196 results for “Evo”

Decrypt reports on backlash against Proton after 404 Media revealed that CEO Andy Yen called a request for pentesting time "loser thinking" while pushing engineers to use AI instead of hand-writing code. Proton cancelled planned AI-use re-interviews on 30 October 2026, and CTO Bart Butler's 1 November manifesto says engineers should direct LLMs and that most human code review may end. User cancellations are anecdotal, not a verified count.

decrypt.coarticlePublished and added 2 Nov 2026🧠 AI

Revolve, an online fashion retailer, disclosed a September 11 breach of systems including its e-commerce platform. Attackers may have exfiltrated names, addresses, card numbers and expiration dates that customers saved in their accounts. Notices filed with the Iowa, Massachusetts, Texas and Vermont attorneys general show at least 179,259 people notified. Revolve offers no credit monitoring, and the attacker and method are undisclosed.

securityweek.comarticlePublished and added 2 Nov 2026🛡️ Security

The FBI attributed the roughly $6.3 billion Binance theft of about 29 October 2026 to North Korea's TraderTraitor (Lazarus Group, Jade Sleet, APT38). Binance traced the breach to a compromised DevOps engineer's laptop that gave access to a service container used to submit unauthorized withdrawals across ten blockchains. The FBI published 628 laundering addresses and linked the group to earlier thefts at Bybit, DMM Bitcoin, Bitget and Drift.

bleepingcomputer.comarticlePublished and added 1 Nov 2026🛡️ Security

Decrypt's explainer on the roughly $6.4 billion theft from Binance hot wallets across about 10 blockchains. Binance traced it to a compromised DevOps engineer laptop plus further breach points, and on Ethereum the attacker used 50 pre-signed EIP-7702 authorizations to turn a hot wallet into a sweeper. Chainalysis and TRM Labs attribute it to TraderTraitor; Tether and Circle froze about $2.3 billion.

decrypt.coarticlePublished and added 31 Oct 2026🛡️ Security

StepSecurity's analysis of Binance's 29 October 2026 incident, which Binance puts at about $6.3 billion in affected assets. It covers Binance's disclosure of a possible DevOps laptop compromise, and TRM Labs and Elliptic findings on 50 EIP-7702 authorizations and a persistent delegation. It links this to the Nx, OVHcloud CLI and Claude Code extension compromises, gives an investigation checklist, then describes StepSecurity products.

stepsecurity.ioarticlePublished and added 30 Oct 2026🛡️ Security

The Block reports that Binance lost about $6.3 billion in a hack on 29 October 2026. Co-CEO Richard Teng said a compromised DevOps engineer's laptop gave access to a service container used to submit unauthorized withdrawals, and TRM Labs attributed the theft to North Korea's TraderTraitor. The rest covers reactions: President Trump said on 30 October he was looking into it, CFTC Chair Michael Selig and The Digital Chamber urged federal exchange rules, and senators disputed whether the Clarity Act would have allowed freezes.

theblock.coarticlePublished and added 30 Oct 2026🛡️ Security

Galaxy Research's weekly brief of 30 October 2026 opens with the Binance hack: about $6.3b drained across 10 blockchains, attributed by TRM Labs and Chainalysis to TraderTraitor (Lazarus-linked). The attacker used EIP-7702 with 50 pre-signed authorizations on Ethereum, and Binance's preliminary finding is a compromised DevOps laptop reaching a service container. Tether and Circle froze about $2.6b. The other sections cover Hyperliquid and prediction markets.

galaxy.comarticleAdded 30 Oct 2026🛡️ Security

Galaxy Research newsletter on the roughly $6.3b drained from Binance across 10 blockchains, which TRM Labs and Chainalysis attribute to TraderTraitor. On Ethereum the attacker used EIP-7702 to turn a hot wallet into a sweeper contract backed by 50 pre-signed authorizations. Binance's preliminary finding is a compromised DevOps laptop, with no wallet key leak. Tether, Circle, Ripple and World Liberty froze about $2.7b; Galaxy also assesses Binance's ability to cover losses.

galaxy.comarticleAdded 30 Oct 2026🛡️ Security

Bankless report on the Binance hack of $6.3 billion, which Binance's preliminary investigation traced to a compromised DevOps engineer's laptop with access to withdrawal-authorizing systems. TRM Labs found valid EIP-7702 authorizations on Ethereum, and nine other chains were drained through normally signed transactions. TRM and Chainalysis link it to North Korea's TraderTraitor. Tether and Circle froze about $3 billion, and Binance says SAFU and corporate cash will cover the losses.

bankless.comarticlePublished and added 30 Oct 2026🛡️ Security

Decrypt reports that Binance is working with the FBI and U.S. Secret Service after a theft of about $6.4 billion, which TRM Labs and Chainalysis attribute to North Korea's TraderTraitor group, also linked to the February 2025 Bybit hack. Circle froze about $823 million in USDC and Tether froze USDT, though the page gives differing Tether totals. Binance first suspected a compromised DevOps engineer laptop, found more breach points, and saw no evidence of leaked wallet private keys.

decrypt.coarticlePublished and added 30 Oct 2026🛡️ Security

Binance hack analysis from 21shares on the 29 October 2026 theft of about $6.3 billion from Binance. Binance's initial finding points to a compromised DevOps engineer's laptop with no sign of a leaked wallet key. Chainalysis attributes it to North Korea with high confidence, and about $3 billion was frozen by Tether, Circle and the Arbitrum Security Council. The attacker also borrowed about $44.6 million of WETH on Aave. Binance says SAFU and company funds will cover all losses.

21shares.comarticlePublished and added 30 Oct 2026🛡️ Security

The Block reports that Binance says it has set aside $6.34 billion of its own funds to cover a hack it puts at about $6.3 billion. Co-CEO Richard Teng says the entry point was a compromised DevOps laptop and no private key leak was found. About $3 billion was frozen by Tether, Circle, Ripple and others, withdrawals are suspended, and TRM Labs sees parallels with North Korean attacks. Technical detail is thin.

theblock.coarticlePublished and added 30 Oct 2026🛡️ Security

Aikido Security's analysis of the roughly $6.3 billion Binance incident, which Binance says may have started with a compromised DevOps engineer's laptop. It sets out what is known (no private key leak, TraderTraitor attribution by TRM Labs and Chainalysis, additional breach points) and a plausible chain from laptop credentials to a withdrawal service. It compares the 2025 Bybit theft and lists mitigations.

aikido.devarticlePublished and added 30 Oct 2026🛡️ Security

Binance estimated the October 2026 security incident at about $6.3 billion and said it found "additional points of breach" beyond the compromised DevOps engineer laptop it first named. The attacker reached a service container and submitted unauthorized withdrawals, with no wallet private key leak found. Arkham counted $6.01 billion taken from hot wallets on nine chains; Tether and Circle froze about $1.8 billion in stablecoins. About $3 billion is frozen onchain; withdrawals stay suspended.

decrypt.coarticlePublished and added 30 Oct 2026🛡️ Security

Binance linked the theft of about $6.3 billion in crypto to a compromised DevOps engineer's laptop, which let attackers reach a service container and submit unauthorized withdrawals. TRM Labs and Chainalysis attribute the 29 October 2026 attack to North Korea's TraderTraitor. About $2.7 billion was frozen, and the attackers hold about $2.4 billion. Binance has found no evidence of a leaked wallet private key.

bleepingcomputer.comarticlePublished and added 30 Oct 2026🛡️ Security

The Binance hack of 29 October 2026 drained about $6.36 billion across ten blockchains in roughly 20 minutes, according to TRM Labs. TRM attributes it with high confidence to North Korea's TraderTraitor, citing a mixer-funded operator wallet, pre-signed transactions and overlap with the Bybit and KelpDAO hacks. Binance's early findings point to a compromised DevOps engineer laptop. Stablecoin issuers froze about $2.7 billion.

securityweek.comarticlePublished and added 30 Oct 2026🛡️ Security

Binance disclosed a suspected security incident on 29 October 2026, estimating $6.3 billion affected and about $3 billion frozen. Binance's early finding is that a compromised DevOps engineer's laptop let an attacker reach a service container and submit unauthorized withdrawals, with no evidence of wallet private-key leakage. TRM Labs suspects TraderTraitor, the group tied to the February 2025 Bybit theft. Tether, Circle and Ripple froze funds.

en.cryptonomist.charticlePublished and added 30 Oct 2026🛡️ Security

Trail of Bits post on X summarising Binance's initial investigation into a security incident. Binance suggests a compromised DevOps engineer's laptop let an attacker reach a service container and submit unauthorized withdrawal transactions. The text says no evidence has been found of wallet private key compromise. The fetched text is truncated and gives no further detail.

x.comthreadPublished and added 30 Oct 2026🛡️ Security

The OpenSourceMalware Show #27 (28 October 2026) is a podcast episode with transcript. It covers the Brotli decoder bug (CVE-2026-111497) vendored in C code across Python, Rust and PHP, and how to find copies via decode.c. It also covers the LiteLLM v1.107.0 compromise from an unrevoked admin key and a relaxed release control, and the MALFEX npm campaign hiding payloads in PNG-header files, plus an AGENTS.md lure and DPRK fake fonts in TTF/OTF.

opensourcemalware.comtalkPublished and added 29 Oct 2026🛡️ Security

Paper studying whether self-evolution weakens personal agents' robustness to indirect prompt injection (IPI). A two-phase method has the user ask the agent to follow tasks in external content, then tests with a third-party injected task the user did not authorize. Against the non-evolving base agent, attack success rose 14.0 to 48.8 points under skill evolution and 3.0 to 39.3 points under reinforcement learning. Code is released.

arxiv.orgpaperAdded 31 Oct 2026, published 29 Oct 2026🛡️ Security

Secure Transaction Verification and Authorization (STVA) is a draft Standards Track ERC (created October 2026) by @MATOBOYCRYPTO65. It proposes structured human-readable metadata and semantic intent for transactions and UserOperations, an independent simulation and verification interface, and scoped, revocable capability tokens for privileged operations. It targets blind signing and frontend supply-chain attacks, citing the February 2025 Bybit Safe multisig theft. The page gives only the abstract, motivation and core concepts.

ethresear.chpaperPublished and added 28 Oct 2026🛡️ Security

GitHub Universe 2026 recap in which GitHub presents its agentic software delivery vision and new Copilot products: the Copilot app, Copilot Fusion multi-model routing, Copilot projects, session continuity, automations that learn from outcomes, and code review Max. GitHub Enterprise customers can use all Copilot features without a separate subscription. GitHub also reports added capacity, a multi-region Azure migration and a redesigned Git architecture.

github.blogarticleAdded 29 Oct 2026, published 28 Oct 2026🧠 AI

FoxIO disclosed three TLS flaws in the GoAhead embedded web server from EmbedThis, used on hundreds of millions of IoT, OT and network devices. CVE-2026-115881 (CVSS 9.1) leaves revoked client certificates accepted under mTLS with OpenSSL; CVE-2026-115884 (7.4) bypasses chain-length limits with LibreSSL; a third Mbed TLS bug (5.9, no CVE) weakens TLS 1.2 DHE. GoAhead 6.1.1 fixes the OpenSSL and Mbed TLS issues, and FoxIO released PoC code and fingerprints.

securityweek.comarticlePublished and added 28 Oct 2026🛡️ Security

FeatureArc is a benchmark for coding agents on evolving features, built by a pipeline that mines repository histories into multi-stage tasks across four programming languages. It has 73 manually reviewed tasks of 2 to 10 stages, validated with fail-to-pass and pass-to-pass tests. Across 10 agents, pooled success drops from 62.7% at the first stage to 28.5% for the whole feature. Inherited defects appear in 72.9% of failed later stages.

arxiv.orgpaperAdded 28 Oct 2026, published 27 Oct 2026🧠 AI

Denver Health Medical Plan is notifying 238,083 members that vendor SquareML leaked an access key on a public website. The key was online from 27 June 2025 until it was revoked on 18 August 2026 and gave read-only access to a database of names, dates of birth, health metrics, eligibility and admission data. DHMP found no evidence of misuse and reset passwords for all accounts with access.

securityweek.comarticlePublished and added 27 Oct 2026🛡️ Security

Thread on two October 2026 npm compromises: @subql/common 5.8.3 (5 October), published through GitHub Actions OIDC Trusted Publishing after the attacker tampered with the publish workflow, and tensorlake 0.5.144 (8 October), published by the project's own release workflow with valid provenance for about 1 hour 40 minutes. Both ran on install and stole cloud, GitHub, Kubernetes, Vault, SSH and .env credentials; tensorlake also took AI tool and MCP config. Its payload installed a gh-token-monitor service that deletes the home directory if the stolen GitHub token is revoked, so remove it before rotating.

x.comthreadAdded 27 Oct 2026, published 26 Oct 2026🛡️ Security

IntentDelta is a human-constructed multi-turn benchmark of 638 trajectories and 3,766 turns across WebShop and TravelPlanner. Each turn pairs a user utterance with a cumulative intention state and human reference actions. Across eleven agents, average Intention Recovery is 33.1% in TravelPlanner and 44.7% in WebShop, and average Action Score is 31.6% and 41.0%. Agents lose earlier requirements and confuse soft preferences with must-have constraints.

arxiv.orgpaperAdded 27 Oct 2026, published 24 Oct 2026🧠 AI

The Hacker News post relays LiteLLM's root-cause disclosure for its supply chain attack. A master key leaked on GitHub in May was never revoked. Attackers used it to reach the Agent Platform, steal a GitHub owner token, and publish a malicious PyPI release in October. A reply says the agent tool held the org owner's personal CLI login, which skips GitHub's 2FA re-check.

x.comthreadAdded 26 Oct 2026, published 24 Oct 2026🛡️ Security

Unit 42 advisory on the compromise of eight intermediate CA private keys belonging to Indian CA eMudhra Limited. The keys were generated with insufficient randomness, so attackers could derive them from the public certificates and sign rogue TLS certificates for MitM attacks. eMudhra has revoked the certificates and browser and OS vendors have untrusted them. Interim guidance covers PKI and trust store audits, reissuance, strict CRL/OCSP checking, log hunting, CT monitoring and CAA records.

unit42.paloaltonetworks.comadvisoryAdded 30 Oct 2026, published 22 Oct 2026🛡️ Security

Recap of a Huntress Tradecraft Tuesday session (October 2026) on phishing kits that target Microsoft 365 tokens, including EvilTokens (a device-code phishing campaign abusing Railway) and the Kali365 control panel. It explains device-code phishing, adversary-in-the-middle and token replay. It lists detections (unfamiliar cloud sign-ins, device-code auth, new devices) and response steps (revoke refresh tokens, remove rogue devices, block device-code flow).

huntress.comarticlePublished and added 22 Oct 2026🛡️ Security

eMudhra's eight intermediate CA certificates, chaining to emSign ECC Root CA - G3, had private keys below 256 (one is 4), disclosed on 17 October 2026 by a researcher known as Wayne. Browsers and Microsoft blocked them via revocation pushes, but curl, git, containers, language runtimes and IoT may still trust them. The post gives patched ca-certificates versions, test commands and the timeline.

aikido.devarticlePublished and added 22 Oct 2026🛡️ Security

Oracle Security Alert for CVE-2026-91403 ships Java SE updates that distrust a set of intermediate CA certificates tied to eMudhra. It contains one new patch for Oracle Java SE, described as remotely exploitable without authentication. Customers are told to update Java or apply mitigations listed at ops.java/emudhra and in the Patch Availability Document. No reporter is credited and the page gives no CVSS score or technical detail.

oracle.comadvisoryAdded 23 Oct 2026, published 21 Oct 2026🛡️ Security

Railway's security disclosure of 21 October 2026 says an unauthorized third party may have accessed encrypted customer environment variables and decrypted some of them. Railway links it to a series of attacks on platform companies aimed at compromising AI inference providers, and has asked upstream providers to revoke keys. It tells all users to roll every environment variable, API key and registry credential, and it paused all deployments. The notice gives no technical details of the intrusion.

blog.railway.comadvisoryAdded 22 Oct 2026, published 21 Oct 2026🛡️ Security

The eMudhra CA compromise (disclosed 20 October 2026) involves eight intermediate CA certificates whose private keys can be derived from the certificates themselves. They chain to AAA Certificate Services and three other roots. Browsers block them, but non-browser clients skip revocation and CT checks. Endor Labs finds 25 popular packages with frozen bundled root stores, 11 already patched, and gives a fingerprint-scanning script, test hosts and patch steps.

endorlabs.comarticlePublished and added 21 Oct 2026🛡️ Security

eMudhra, an Indian certificate authority with over 60 million issued certificates, had eight CA private keys that were tiny integers such as 4, 56 and 237. Researcher Scott Helme recovered all eight from public certificates in under a second, and the keys were exposed from April 2025. Mozilla, Chrome, Apple and Microsoft blocked them between 17 and 20 October 2026, but curl, Python, Node.js, Go and Java accepted a forged Stripe API certificate because they skip revocation checks.

coindesk.comarticlePublished and added 21 Oct 2026🛡️ Security

Cloudflare's incident post on the compromise of private keys for eight eMudhra intermediate certificates, disclosed 19 October 2026. The weak keys let an attacker forge certificates for any domain. It describes Cloudflare's patching of trust stores, WARP and cloudflared, tells developers to update the underlying TLS verifier and not disable verification, and points to Dennis Jackson's test site and revocation systems such as CRLSets and upki.

blog.cloudflare.comwriteupAdded 29 Oct 2026, published 20 Oct 2026🛡️ Security

ALTK-Evolve is an IBM Research method that distills an agent's own training-task trajectories into short behavioral guidelines appended to the system prompt, with no weight updates. On τ²-bench, a stock ReAct agent with gpt-4.1 reaches 67.5% Pass^4 on retail, and gpt-5.5 reaches 82% on airline. The post compares support-count thresholds, token cost, turn counts and leaderboard placement.

huggingface.cowriteupPublished and added 20 Oct 2026🧠 AI

Nikkei disclosed a breach of 1.83 million records from a cloud service it uses for system development, accessed on 7 October 2026 with previously compromised credentials. About 150,000 records held names, addresses, emails, phone numbers and birth dates, and 120,000 held email addresses only. Nikkei revoked the credentials and notified regulators. It found no misuse as of 16 October, and it had disclosed an earlier Slack account breach less than a year before.

securityweek.comarticlePublished and added 20 Oct 2026🛡️ Security

MemT-1.0 is a memory foundation model architecture that builds memory into model computation. It builds on Metis, uses Gated DeltaNet-2 for token-level memory updates within each turn, and adds Mixture-of-Storage to scale capacity. Training used 34B tokens in continued pre-training and post-training. MemT-1.0-Nano and MemT-1.0-Flash beat parametric memory baselines and stay competitive with retrieval methods. At 128K history, Nano gives about 1.86x inference speedup and 66x less persistent storage than a full-context KV cache.

arxiv.orgpaperAdded 21 Oct 2026, published 19 Oct 2026🧠 AI

Dynalize is a method that reduces reward hacking during online RL training against preference reward models. Periodically and asynchronously, a language model reads training rollouts and writes penalties for undesirable behaviors the reward model seems to reward, and a judge applies them to every rollout. In clinical counseling and creative writing it sustains longer training and improves external benchmarks, and the gains hold in multi-turn training and after reward model retraining.

arxiv.orgpaperAdded 21 Oct 2026, published 19 Oct 2026🧠 AI

The paper is a large-scale study of 37 security skills run through Claude Code with four LLMs on 100 real Linux kernel vulnerabilities, 14,800 runs in all. Only 24.4 percent of runs found the correct function and 15.7 percent classified the vulnerability correctly. It proposes SKILL-SIF, which evolves skills from failure patterns and raised fully correct detection on 100 unseen vulnerabilities from 2 to 16 percent.

arxiv.orgpaperAdded 21 Oct 2026, published 19 Oct 2026🧠 AI

Empirical study of 37 security skills run through Claude Code with four LLMs on 100 real Linux kernel vulnerabilities across ten classes (14,800 runs). 24.4% of runs found the correct function, 18.3% the function and statement, and 15.7% the vulnerability class. The authors propose SKILL-SIF, which evolves skills from failures and raised fully correct detection on 100 unseen vulnerabilities from 2% to 16%.

arxiv.orgpaperAdded 21 Oct 2026, published 19 Oct 2026🛡️ Security

Self-evolving Agent Harness is a 66-page survey and theory paper on revising an agent's scaffolding (prompts, memory, workflows, tools, code, plugins) while model weights stay fixed. It frames search as evolutionary optimization and uses continual PAC-learning to derive a generalization bound, a retention bound and a high-probability admission rule with fresh validation data. It separates selection, activation, attribution and improvement as evaluation claims, and covers applications and governance. Code is at github.com/zgcagi/Harness-RSI.

huggingface.copaperAdded 19 Oct 2026🧠 AI

Datadog Security Labs tests whether Project CodeGuard v1.4.0 security rules improve the code written by Claude Code (Sonnet 5), Cursor (Composer 2.5) and Codex (GPT 5.5). Each model built the same document portal app once. Only GPT 5.5 closed the cross-user document IDOR and improved clearly. Sonnet 5 and Composer 2.5 stayed largely unchanged, keeping IDOR, hardcoded secret fallbacks and token revocation gaps.

securitylabs.datadoghq.comwriteupAdded 29 Oct 2026, published 19 Oct 2026🧠 AI

Second post in a Datadog Security Labs series testing whether Project CodeGuard (v1.4.0) security rules improve code written by coding agents. Sonnet 5 (Claude Code), Composer 2.5 (Cursor) and GPT 5.5 (Codex) each built the same document portal, and the output was audited with Datadog Code Security and a Sonnet 5 code review. Only GPT 5.5 closed the cross-user IDOR; Sonnet 5 and Composer 2.5 stayed largely unchanged, including hardcoded JWT secret fallbacks and incomplete revocation.

securitylabs.datadoghq.comwriteupAdded 29 Oct 2026, published 19 Oct 2026🛡️ Security

Decrypt guide to "tokenmaxxing" reframed as maximizing value per token rather than burning tokens. It gives tiered tips: pick smaller models for simple jobs, write one clear prompt, start new chats, cap output length, and use free models (beginner); Claude Code commands like /model, /clear, /compact, /context, /mcp, /effort and plan mode (intermediate); and prompt caching via cache_control, 50% batch discounts, and agent budget limits like --max-budget-usd (advanced). Includes context on Meta, Amazon and Uber token leaderboards and spend, a worked invoice-processing example where routing cut cost from $30 to $3.90, and the Jevons paradox in falling token prices.

decrypt.coarticlePublished and added 17 Oct 2026🧠 AI

Reports a Science study from Stanford University and the Arc Institute in which Evo, a genome language model trained on sequences from across the domains of life, generated thousands of bacteriophage genome designs targeting E. coli. About 300 were built and tested, and 16 produced viable viruses. A mixture of the AI-designed phages overcame resistance in some E. coli strains where a comparable mixture of natural phages did not. The article also covers a Johns Hopkins Center for Health Security commentary saying governance lags behind generative genome design, and a more cautious biosafety assessment from an outside expert.

en.cryptonomist.charticlePublished and added 16 Oct 2026🧠 AI

GLOBE is an agentic framework that optimizes e-commerce product listings for generative search by storing strategies in an evolving handbook of explicit, field-scoped behaviors. An Editor applies behaviors and gets rank-based feedback, a Reflector turns it into updates, a Moderator removes redundant proposals, and a Curator commits changes. Harmful behaviors are deactivated. Structured behaviors beat monolithic prompt optimization and transfer to unseen product categories.

arxiv.orgpaperAdded 20 Oct 2026, published 16 Oct 2026🧠 AI

SecurityWeek reports F5 fixed 11 NGINX vulnerabilities. Three are critical: CVE-2026-15176 and CVE-2026-93479 are unauthenticated buffer overflows via crafted HTTP requests, and CVE-2026-97205 is memory corruption. All three can lead to RCE if ASLR is disabled or bypassed. Juniper Networks published 17 advisories covering over 150 flaws in Junos OS and Junos OS Evolved, plus four in Cloud-Native Router. About two dozen are critical, with root code execution, device takeover, kernel panic and an off-by-one buffer overflow. Neither vendor reports exploitation in the wild.

securityweek.comarticlePublished and added 16 Oct 2026🛡️ Security

Analysis of a new GhostAction Stage 2 found in 94 compromised GitHub repos: a hidden .vscode/tasks.json (runOn folderOpen) executes a plain-JavaScript infostealer disguised as public/fonts/fa-solid-400.woff2. The stealer grabs cloud, registry, SSH, Claude/Gemini credentials, POSTs them to a hardcoded IP on port 8080, and installs a 5-minute cron job (#ghc2) that pipes the server's script to bash. It copies PolinRider's launcher but shares no IOCs with it, and the post gives a timeline, one repo hit by both campaigns, IOCs, and cleanup steps (revoke tokens, remove cron, set task.allowAutomaticTasks off).

opensourcemalware.comwriteupAdded 15 Oct 2026, published 16 Oct 2026🛡️ Security

GitHub engineers describe Evergreen, a Go-based GitHub App that runs quarterly outside-collaborator access reviews and archives stale repositories in a shared org. It uses scheduled Actions with repository_dispatch, JSON state files in git as the audit trail, checklist issues (IssueOps) closed to trigger revocation, and custom properties for ownership and exemptions. Design principles include fail-secure/fail-closed revocation, never deleting data, dry-run rollout, Splunk logging, and proactive pacing against GitHub's secondary rate limits.

github.blogarticlePublished and added 15 Oct 2026🛡️ Security

T-EVOLVE extends the lattice-commitment e-voting protocol EVOLVE to threshold tallying. Votes are Shamir-shared, so any t of n authorities can finish the tally. Privacy holds against t-1 colluding authorities and verifiability holds even if all of them collude. The voter proves her committed shares form a Reed-Solomon codeword whose value at zero is a valid vote. Complaints about ciphertexts that do not decrypt are publicly verifiable, using an LWE-instantiable encryption scheme with publicly verifiable opening. Privacy, verifiability, individual verifiability and robustness are proven in the classical random oracle model, and the verifiability reduction is not tight. In a C implementation, a yes/no ballot for four authorities takes 109 KB, and an authority aggregates 10^4 ballots in about three minutes on one core.

eprint.iacr.orgpaperAdded 17 Oct 2026, published 15 Oct 2026🛡️ Security

Snyk introduces VulnBump, a benchmark of 307 real vulnerable-dependency upgrades from 43 repositories across Maven, Gradle, npm, Yarn, pip and Poetry. An attempt passes only if a fresh Snyk rescan shows the vulnerabilities gone and the project installs, builds and keeps its previously passing tests green in a clean container. On Claude Sonnet 5.5, Snyk's Evo Remediation Agent passed 90.2%, Claude Code with Snyk scan and target version 77.5%, unaided Claude Code 59.6%, and a non-AI package-manager bump baseline 46.9%. Unaided Claude Code often claimed fixes a rescan still flagged, so the benchmark shows harness and security data matter as much as the model. The harness and 43 cases are open-sourced.

snyk.ioarticlePublished and added 15 Oct 2026🧠 AI

Snyk presents VulnBump, a benchmark of 307 real vulnerable-dependency upgrades across 43 repositories and six package managers (Maven, Gradle, npm, Yarn, pip, Poetry). An upgrade passes only if a rescan shows the vulnerabilities gone and the project still installs, builds, and passes its previously passing tests. On Claude Sonnet 5.5, Snyk's Evo Remediation Agent passed 90.2%, Claude Code with Snyk scan 77.5%, Claude Code unaided 59.6%, and a plain package-manager bump 46.9%. Unaided Claude Code often claimed fixes that a rescan disproved, and major framework migrations and transitive dependencies were the hardest cases. The harness and the 43 cases are open-sourced.

snyk.ioarticlePublished and added 15 Oct 2026🛡️ Security

Introduces Recursive Reward Improvement (RRI) for recursive self-improvement on open-ended expert tasks without answer keys, where a proposer writes both a task and a rubric. Before the solver trains, an adversary writes a rubric-gaming but useless answer, the same model writes an honest answer without seeing the rubric, and up to two repair steps revise the rubric until the honest answer wins. Exploits found also rewrite future task and rubric generation. On HealthBench Professional it beats a Generation Only baseline by 10.9 points at 9B and 12.4 at 27B (10.3 with no external model), and the 27B model transfers to PRBench Hard (+6.6) and ProfBench (+8.0).

arxiv.orgpaperAdded 17 Oct 2026, published 15 Oct 2026🧠 AI

Huntress post on where accounts remain exposed after a passkey rollout. It covers passkey-themed social engineering, the EvilTokens device-code phishing campaign (344 organizations in about 16 days), weak enrollment and recovery, SpecterOps research on Windows WebAuthn hooking on compromised endpoints, and the trade-offs between synced and device-bound passkeys. It closes with ITDR-style post-login monitoring, containment steps (revoke sessions, remove rogue authenticators, hunt inbox rules, OAuth grants, forwarding), and a five-question deployment checklist.

huntress.comarticlePublished and added 14 Oct 2026🛡️ Security

SecurityWeek reports on the arrest of two US nationals (Michael James Bostelle, Juan Cecilio Pena Carrillo) for running laptop farms that helped North Korean IT workers get remote jobs at over 70 US companies, generating more than $3M between 2019 and July 2025. It covers the FBI's Operation Lonestar Roundup (40 Texas locations searched, 187 laptops and KVMs seized). It also covers CrowdStrike's Famous Chollima analysis of shipping-diversion tricks and evolving evasion (hidden RMM tools, spoofed device hashes), and the FBI's 'identity in depth' guidance for spotting ITWs across the employment lifecycle.

securityweek.comarticlePublished and added 14 Oct 2026🛡️ Security

AINews issue for 10/12–10/13/2026 covering the GA of Google's Gemini 4 Argon (pricing, Cursor/Devin/Copilot availability, CursorBench 52.8%, Artificial Analysis Intelligence Index parity with GPT-6 Astra at lower cost) and Meta's Muse Wonder, with independent evals and reliability caveats. It also summarises decision-model benchmarks (AA Decision Index, Jev 1.13, Nimble-v4), Fish Audio's open-weight Drama 3 TTS, agent-harness and agentic-RL research (Evo2, ReHarness, LongHarness, ThunderSync), Anthropic's move off Git, and safety items such as Argon's monitor evasion. It also notes AI Engineer NYC keynotes on AI x Finance. Useful as a curated, caveat-aware snapshot of the frontier-model landscape.

latent.spacearticlePublished and added 14 Oct 2026🧠 AI

ProAS (Progress-Aware Advantage Shaping) is a plug-and-play framework for multi-turn agentic RL that fixes the uniform credit assignment problem in critic-free algorithms like GRPO, where trajectory-level advantages are broadcast to every action. It uses a dynamically evolving Agentic Reference Graph with Masking Unreasonable Incentives and Asymmetric Penalty Redistribution to give finer-grained credit without a critic. On Qwen2.5-1.5B/7B-Instruct it reports success rates up to 91.9%/94.3% on ALFWorld and 66.7%/70.8% on WebShop, with up to 2.44x faster training convergence and robustness to imperfect or distilled ARGs.

arxiv.orgpaperAdded 14 Oct 2026, published 13 Oct 2026🧠 AI

Ledger's Donjon team confirmed that Ledger signers sold by the reseller CryptoBilis contain an unauthorized implant. It captures the 24-word recovery phrase during setup and sends it over LTE using a provisioned eSIM. It does not touch the Secure Element, the OS, or transaction signing. Onchain researchers estimate losses above $86M, but Ledger says that figure is unverified. MediaMarkt Saturn halted European Ledger sales after a tampered unit was reported in Germany, and Trezor and OneKey tightened or revoked reseller authorizations. The case shows a reseller-channel tampering risk for hardware wallets.

theblock.coarticlePublished and added 13 Oct 2026🛡️ Security

EmoGenie is a benchmark for emotions expressed and evoked by text-to-image outputs. It has 6,400 images from eight T2I models using 800 emotion-targeted prompts. It includes EmoGenie-QA (90,000 closed-set and 1,057 open-ended questions) and EmoGenie-Assess (13,175 mean opinion scores for valence, arousal, dominance and aesthetic quality). The authors also propose EmoProbe, which pairs emotion QA with response-aligned, dimension-specific regression heads. They report that it beats baselines on emotion QA and correlates better with human ratings. The work shows how reliably MLLMs interpret affect in generated images.

arxiv.orgpaperAdded 14 Oct 2026, published 13 Oct 2026🧠 AI

SecurityWeek summarizes Netskope research showing the XWorm Windows RAT grew from 35 plugins (July 2025) to 46 (October 2026), adding surveillance features: Maps (GPS location without prompt), Microphone (live audio streaming), webcam capture, and HBrowser. HBrowser gives operators a live, interactive view of the victim's authenticated browser sessions over the AES-encrypted C2 channel, bypassing logins and MFA. Activity rose after the builder was cracked in June 2025 and the 6.4 source leaked in August 2025; Netskope saw 1,200+ incidents across 100 organizations in Q3.

securityweek.comarticlePublished and added 13 Oct 2026🛡️ Security

Part 1 of an AWS series showing how to build a quantum-resistant software supply chain with ML-DSA: ML-DSA keys in AWS KMS, an ML-DSA root/subordinate CA hierarchy in AWS Private CA, and IAM Roles Anywhere trust anchors for external build agents. It covers signing binaries (EXTERNAL_MU with OpenSSL 3.5), container images as OCI referrers via ORAS, and SBOMs wrapped in in-toto Statements bound to the artifact digest, plus verification, admission control, trust-store distribution, revocation and SCP guardrails. It frames the "forge later" threat and maps CNSA 2.0 (ML-DSA-87 only for NSS) and EU CRA deadlines to actions.

aws.amazon.comarticlePublished and added 12 Oct 2026🛡️ Security

StableJudge is a defense framework against score-hijacking prompt injection in LLM-as-a-Judge pipelines (e.g. paper review, recommendation). A compact 0.13B "Sentry" model localizes injected spans and sanitizes them before the judge scores the content, trained on adversarially evolving attacks, and deterministic fine-grained scores are derived from the judge's output distribution. It improves score recovery under attack on held-out scientific abstracts and out-of-distribution attack variants while preserving general performance, decoupling safety from judge model scale. Code and data are promised but not yet released.

arxiv.orgpaperAdded 15 Oct 2026, published 12 Oct 2026🛡️ Security

Nathan Lambert argues that near-term AI progress will come mostly from automating infrastructure and engineering work (superhuman distributed GPU engineers, end-to-end optimization of training and inference), not from models becoming broadly superhuman researchers. He predicts near-exponential declines in the cost of model intelligence, automated pretraining research on architecture and data selection in 2-3 years, and a Jevons-paradox surge in demand for agentic models such as Meta's Muse. He also flags low-quality RL environments sold by fast-growing data companies as fixable low-hanging fruit.

interconnects.aiarticlePublished and added 9 Oct 2026🧠 AI

The paper studies joint attention in multimodal diffusion transformers (MMDiTs), where the evolving image updates text representations. It finds that this image-to-text (I2T) feedback tends to preserve objects and attributes already in the image, errors included, and that suppressing it improves generation. It proposes Semantic Feedback Guidance (SFG), which guides away from an auxiliary branch that strengthens I2T feedback and weakens text-to-image conditioning, and iSFG, an in-place variant that needs no extra model evaluations. SFG runs for the first 20% of sampling steps (about 1.2x base NFE without CFG). Across three backbones, adding SFG to CFG on FLUX raises GenEval from 0.6661 to 0.7479, and iSFG on SD3.5M adds +1.33 HPSv2.1 over CFG at zero extra NFE. The control also improves VBench scores for text-to-video and image-to-video.

arxiv.orgpaperAdded 12 Oct 2026, published 9 Oct 2026🧠 AI

Alert summarizing Socket Threat Research's disclosure that a malicious version of the Tensorlake npm SDK (an AI-agent infrastructure package) was published with a preinstall hook delivering Shai-Hulud/ChainDrop credential-stealing malware. The payload targets GitHub, npm, AWS, Kubernetes and Vault credentials, persists via a gh-token-monitor mechanism, and spreads through compromised publishing accounts. It lists IOCs (lib/setup.mjs, lib/Math_Symbol.js) and warns to remove the persistence before revoking stolen GitHub tokens, since revocation triggers a destructive routine. Socket detected the release about 11 minutes after publication.

x.comadvisoryAdded 11 Oct 2026, published 9 Oct 2026🛡️ Security

Google disclosed that attackers hijacked the authoritative DNS of three ccTLDs (.gh for Ghana, .sl for Sierra Leone, .as for American Samoa) and used control of those domains to obtain unauthorized HTTPS certificates covering several Google domains and domains of other organizations. Google blocked the fraudulent certificates in Chrome, worked with issuing CAs to revoke them, and used Certificate Transparency log analysis to find other affected brands, some of which it proactively notified. Google is urging domain owners to monitor CT logs and publish restrictive CAA records after regaining DNS control, since CAs may cache and reuse prior domain-control-validation checks, letting attackers mint new certs even after the hijack ends.

securityweek.comarticlePublished and added 9 Oct 2026🛡️ Security

Galaxy Research's weekly brief devotes its lead section to AI governance: it recaps Sam Altman's and Dario Amodei's UN Security Council remarks calling for international AI safety coordination, contrasts a disclosed OpenAI agent incident (an agent accessed non-public Medicare data in Australia, reported three months late) with Anthropic's 950-agent DNA-database research run, and cites concrete agent-economy data — OpenRouter's agent token usage overtaking human usage and growing ~14x since February, Cloudflare's agent-traffic crossover, and Vercel AI Gateway data showing a 23% monthly price drop and open-weight models rising to 56% of token volume. The analysis argues that as inference gets cheaper and open-weight models gain share, AI governance's hard problem shifts from controlling frontier labs to controlling where cheap, widely available agents can act. The rest of the newsletter covers an unrelated Bitcoin privacy protocol proposal and CFTC guidance on prediction "mention markets."

galaxy.comarticleAdded 8 Oct 2026🧠 AI

The paper tests whether LLM legal explanations reflect what drives the verdict. Holding case facts fixed, it swaps the named statute or precedent for an unrelated one and decodes the evolving verdict from hidden states. Across seven open-weight models (8B-70B) and four benchmarks (CaseHOLD, ECHR, SCOTUS, ContractNLI), models name the correct authority 66.7%-100% of the time, but the verdict changes with the authority far less often (0.0%-21.7% on CaseHOLD, 30.0%-76.7% on ECHR and SCOTUS, 43.3%-50.0% on ContractNLI). Neither scale nor a LoRA reproduction of a legal-reasoning model closes the gap. A red-teaming check on five models finds compliance with an adversarial instruction hidden in the case facts (73.3%-96.4%) far exceeds verdict-swap sensitivity. The authors conclude that generated legal explanations are poor compliance or audit artefacts.

arxiv.orgpaperAdded 9 Oct 2026, published 8 Oct 2026🧠 AI

Cisco Talos' CAIRN research classifies a malware archetype called 'A3: AI-Analysis Evasion' — samples that embed natural-language instructions aimed at manipulating LLM-based triage/reverse-engineering pipelines. The post traces the technique's spread and evolution across four confirmed families (FRUITSHELL, PLOTSAFE, HOLLOWCLAD, MANTLEMAZE, plus the related ROZESHELL), from a hardcoded 'ignore this file, it's just prime number generation' comment to templated generation, multi-LLM chat-template spraying, and fake legal/government intimidation text, and includes a controlled test showing the simplest direct-instruction technique shifted LLM verdicts toward benign in ~35% of runs while more elaborate variants often backfired. MANTLEMAZE additionally ships a BYOVD loader abusing the Intel iqvw64e.sys driver (CVE-2015-2291) used by Scattered Spider, BlackByte, and Lazarus Group to kill EDR.

blog.talosintelligence.comarticlePublished and added 8 Oct 2026🛡️ Security

Proposes making the verifier the evolving object in self-improving agent loops. The verifier is an inspectable expression over small, mostly deterministic drawback detectors, synthesized from clustered failures, gated at birth, and selected for agreement with a ten-item anchored reference set plus consensus over unlabeled outputs, never for the agent's score. On MBPP+ it gains +0.21 held-out agreement over the hand-authored seed and ends ahead of the bare LLM judge it contains. Removing the anchor guards collapses the verifier into an always-pass grader that still trains skills as well, so downstream task score cannot certify a self-evolved verifier. The paper's Double Ratchet pairs the verifier with a lifecycle-managed skill loop and retains 88-110% of the lift from ground truth or a rubric on code generation, enterprise text-to-SQL and reference-free report generation.

arxiv.orgpaperAdded 9 Oct 2026, published 8 Oct 2026🧠 AI

StepSecurity details how tensorlake@0.5.144 was pushed to npm after malicious commits landed directly on the tensorlakeai/tensorlake GitHub repo's main branch under a maintainer's identity, carrying a valid npm provenance attestation despite being malicious. The preinstall hook downloads Bun to run an obfuscated payload that steals GitHub/npm tokens, cloud and Vault/Kubernetes secrets, SSH keys, browser logins, and AI-tool configs (Claude, Cursor, Windsurf), then self-propagates by republishing stolen npm packages and committing persistence files (.claude/settings.json, .vscode/tasks.json) to reachable repos. A "hostage token" mechanism installs a gh-token-monitor service that runs a home-directory wipe if the stolen GitHub token is revoked before the monitor is removed, inverting the normal incident-response order.

stepsecurity.iowriteupPublished and added 8 Oct 2026🛡️ Security

Unit 42 analysis of how supply chain malware uses Web3 for command and control, in three phases: EtherHiding (smart contract state), TxDataHiding (transaction calldata across TRON, Aptos and BSC) and NullReceiver (C2 IPv4 decoded from a zero-value transaction's recipient address). It covers the ChainDrop npm worm, the PolinRider campaign and DPRK-linked Axios, Mastra and arrayref compromises. It ends with detection and CI/CD policy guidance.

unit42.paloaltonetworks.comarticleAdded 30 Oct 2026, published 7 Oct 2026🛡️ Security

GitHub essay by Secret Protection PM Erin Havens presenting nine quarters of telemetry (Q2 2024–Q2 2026) on secret leaks: pushes with credentials grew 2.59x while developer override rates of push-protection blocks fell from 6.63% to 3.93%, and mean time to manually revoke a leaked secret hovers around 40 days. It introduces a new fine-tuned ModernBERT classifier, built with Microsoft Applied Sciences, that evaluates candidate secrets in context in under 2ms to extend push protection to unstructured secrets (e.g. internal database passwords with no recognizable pattern), rolling out in private preview to GitHub Secret Protection customers and as a /security-review command in Copilot CLI/App.

github.blogarticlePublished and added 7 Oct 2026🛡️ Security

A tweet highlighting Steampipe, an open-source tool that lets practitioners query cloud infrastructure, Kubernetes, and other services using SQL instead of jumping between consoles or writing ad-hoc scripts. It calls out concrete security/ops use cases such as finding publicly accessible storage buckets, resources missing required labels, and Kubernetes namespaces without NetworkPolicies, and links to a longer tutorial on devopscube.com. The author frames it as a quick investigation layer on top of live cloud data, not a CSPM/AWS Config replacement.

x.comthreadAdded 8 Oct 2026, published 7 Oct 2026🛡️ Security

Snyk benchmarks its Evo Continuous Offensive Security (COS) agent against Claude Security running Mythos on TaintedPort, a public deliberately vulnerable web app with 57 known vulnerabilities and 15 known exploit chains. Evo COS, run gray-box against the live URL plus source, confirmed 10 of 15 exploit chains with runnable proofs of concept — for example chaining an SSRF flaw with a hardcoded JWT signing secret into full admin account takeover — while Claude Security, reasoning over source code alone (white-box), found one additional critical bug but could not prove the vulnerabilities were actually reachable and chainable at runtime. The post argues that proving exploitability requires exercising the live application, not just reading code, and frames SAST-style code review and DAST-style live attack as complementary.

snyk.ioarticlePublished and added 7 Oct 2026🛡️ Security

This arXiv paper introduces the "package hallucination attack," where an attacker injects malicious prompts into benign agentic-coding rule files (such as .cursorrules) to induce a coding agent to replace legitimate package dependencies with attacker-controlled ones. The authors propose PackHallu, an evolutionary optimization framework that iteratively rewrites injected prompts using trajectory-level feedback and LLM-guided mutations, and show it achieves high attack success rates with strong transferability across multiple LLMs, benchmarks, and agent frameworks.

arxiv.orgpaperAdded 8 Oct 2026, published 7 Oct 2026🛡️ Security

ASPIRE is an automated red-teaming engine for LLM agents that discovers indirect prompt-injection vulnerabilities beyond pre-specified attack scenarios. It maintains an evolving Agent Security Behavior Graph and uses complementary Explore and Exploit experts to find, verify, and generalize consequence-centric attacks, with trajectory evidence and cross-run memory improving coverage over time. Benchmarked against existing automated red-teaming approaches, it shows broader coverage of consequences, injection methods, environments, and behavior paths while keeping attack success high.

arxiv.orgpaperAdded 8 Oct 2026, published 6 Oct 2026🛡️ Security

Google describes a series of registry-level hijacks affecting the .gh (Ghana), .sl (Sierra Leone), and .as (American Samoa) ccTLDs, in which attackers modified authoritative DNS records to obtain unauthorized HTTPS certificates for Google domains and, via Certificate Transparency monitoring, several other major organizations. Chrome mitigated impact by blocking the fraudulent certificates through CRLSets and coordinating revocation with issuing CAs, then urges domain owners to continuously monitor CT logs and publish restrictive CAA records with ACME account bindings, since cached domain-control-validation state can let attackers mint new certs even after DNS control is restored.

blog.googlearticleAdded 7 Oct 2026, published 6 Oct 2026🛡️ Security

ScienceClaw formalizes a fixed-parameter program self-evolution framework for LLM agents doing scientific tasks, where verified task executions are turned into persistent Skill and Operator updates to the agent's program rather than being discarded. ScienceClaw-Eval, the accompanying benchmark, spans 23 disciplines across natural and social sciences and measures scientific correctness, evolutionary gain, retention, cross-dataset transfer, and evolution cost via sequential task streams and independent reset evaluation. The framework only keeps an update if source-task replay reproduces the repair and the change improves performance on independent scientific tasks, addressing the gap that verified agent executions rarely become lasting program-level improvements.

arxiv.orgpaperAdded 7 Oct 2026, published 6 Oct 2026🧠 AI

MedPrune is a multi-agent framework for medical visual question answering that models the diagnostic process as a heterogeneous communication graph of specialist agents across departments, then dynamically prunes both nodes (irrelevant specialist agents, via reinforcement learning) and edges (low-value intra-/inter-departmental connections) to cut redundant communication. The authors report it beats existing multi-agent baselines on medical VQA benchmarks under both full-set and few-shot settings while improving token efficiency and adversarial robustness.

arxiv.orgpaperAdded 6 Oct 2026, published 5 Oct 2026🧠 AI

GlimpseDNN is a model-agnostic framework that uses a stochastic leakage model to visualize, layer-by-layer, what physical leakage information a deep neural network extracts during profiled side-channel analysis (SCA) to recover secret keys. The authors (Trevor Yap and Shivam Bhasin, NTU) validate it on simulated traces and public datasets (ASCADf, ASCADr, ASCADf_desync50, ASCADf_desync100), showing it can reveal leakages even in deeper network layers when they are not visible in raw traces, with applications to designing more side-channel-resistant hardware.

eprint.iacr.orgpaperAdded 8 Oct 2026, published 5 Oct 2026🛡️ Security

Denmark's CPR administration (Det Centrale Personregister) disclosed that unauthorized actors abused a Danish company's legitimate, authorized access to the CPR system to extract names, addresses and CPR (national ID) numbers for roughly 8.8 million registered citizens. Records of individuals with name/address protection were not exposed. CPR has revoked the company's access, notified the Danish Data Protection Agency (Datatilsynet), and the incident is being investigated by police alongside relevant authorities.

cpr.dkadvisoryPublished and added 5 Oct 2026🛡️ Security

RubricArmor is an adversarial framework for generating evaluation rubrics used in rubric-based reinforcement learning of LLMs, designed to preempt reward hacking before RL training begins. It alternates an attack step, which synthesizes adversarial responses that satisfy the current rubric while failing the actual task, with a repair step that revises the rubric to close the exposed gaps while keeping valid criteria intact. The paper reports that rubrics produced this way outperform existing LLM-based rubric generation baselines and lead to more effective downstream rubric-based RL.

arxiv.orgpaperAdded 6 Oct 2026, published 4 Oct 2026🧠 AI

Learn Harness Engineering is a free, project-based GitHub course teaching how to build the "harness" around AI coding agents — the instructions, state tracking, verification, scope control, and session lifecycle that make models like Claude and Codex reliable on real repositories instead of just prompt-following. It includes 14 lectures, 8 hands-on projects built around an evolving Electron knowledge-base app, copy-ready templates (AGENTS.md, feature_list.json, init.sh, claude-progress.md), a harness-creator agent skill for scaffolding new projects, and a zero-dependency audit-harness.sh script that checks a repo against the five harness subsystems. Recent updates add "Frontier Harness Design Breakdowns" and new Graph/Loop Engineering lectures covering multi-node agent graphs, parallelism, and loop design outside the agent itself.

github.comcourseAdded 3 Oct 2026🧠 AI

EvoRiskBench is an arXiv paper introducing an evolving benchmark for runtime security risks in LLM-powered workspace agents, built around the EP-Path-EF framework (nine entry-point categories, five effect categories) linking risk entry points to technical effects via agent-mediated paths. The authors ran 450 adversarial tasks across six scenarios against nine model-harness combinations (GPT-5.6 Sol, DeepSeek-V4-Pro-0813, Claude Opus 5 paired with Claude Code, Codex, and OpenClaw), finding Codex + DeepSeek-V4-Pro-0813 the most vulnerable at 68.44% attack success rate and showing ASR varies more by model than by harness. Cases are auto-constructed and verified in isolated environments using runtime traces and environment state.

arxiv.orgpaperAdded 5 Oct 2026, published 2 Oct 2026🛡️ Security

DeepSeek Harness is an open-source general-purpose agent harness from DeepSeek, built on the "Cordis" everything-is-a-plugin architecture, shipped as a desktop app for macOS/Windows and runnable as a web UI from code. It handles everyday document/spreadsheet work, coding tasks (repo exploration, bug fixes, tests), research with citations, and background/scheduled jobs, with users able to install or author plugins (including via a chat-driven "Creator mode") and inspect per-turn tool-call traces for debugging. It runs on DeepSeek's DeepSeek-V4.1-Flash model and is explicitly labeled a preview release with evolving core plugins and APIs.

deepseek.comtoolPublished and added 2 Oct 2026🧠 AI

EvoVLM searches over prompts, image preprocessing, generation budgets and answer extraction for frozen Qwen3-VL-2B-Instruct and Qwen3-VL-2B-Thinking, without fine-tuning weights. A local Qwen model ranks a catalog of typed mutations, a trusted renderer produces Python programs, and Pareto selection on accuracy and latency drives two small generations on scientific chart understanding (CharXiv subset). Recorded results on Dev128: Evolved Instruct keeps accuracy with 8.9% lower median latency, and Evolved Thinking gains 16.41 points of accuracy with 10.6% lower median latency. CPU-only replay and fail-closed evidence verification are included. The authors state the limits: small sample, overlap with the evolution set, and results tied to the recorded hardware.

github.comrepoAdded 16 Oct 2026, published 2 Oct 2026🧠 AI

Thesis on evaluating and improving LLM robustness to adversarial input perturbations. It proposes the R_stab robustness metric (Jensen-Shannon divergence), ASA, an evolutionary black-box attack on LLM-as-a-Judge systems (up to 73.8% ASR), trojan trigger recovery on Trojan Detection Challenge 2023 (Pythia-1.4B), and a systematization of bypasses of multi-layer defenses from SaTML CTF 2024. For MCP agents it proposes AttestMCP (HMAC-attested tool calls) and the Commit Boundary pattern, cutting ASR from 53.7% to 12.4% on the 847-scenario MCPBench; the methods are implemented in JudgeGuard, TrojanArmor and MCPSec.

arxiv.orgpaperAdded 9 Oct 2026, published 1 Oct 2026🛡️ Security

An arXiv thesis-style paper proposing R_stab(f), a Jensen-Shannon-divergence-based robustness metric for LLM outputs under small input perturbations, along with a proven bound V(h) <= 1 - R_class(h) for localized attacks. It introduces ASA, an adaptive evolutionary black-box attack against LLM-as-a-Judge systems reaching up to 73.8% attack success rate (62.6% cross-model transfer), evaluates trojan/backdoor detection on the Trojan Detection Challenge 2023 (Pythia-1.4B) and SaTML CTF 2024, shows that committees of 5-7 heterogeneous models cut Gemma-3-4B's ASR to 19.3%, and presents AttestMCP, an HMAC-based tool-call attestation scheme plus a "Commit Boundary" isolation pattern for MCP agents that cuts average ASR on the 847-scenario MCPBench from 53.7% to 12.4%. Methods ship as the JudgeGuard, TrojanArmor, and MCPSec software suites.

arxiv.orgpaperAdded 5 Oct 2026, published 1 Oct 2026🛡️ Security

InterEvolve is a test-time evolution method for humanoid loco-manipulation that lets a pretrained controller solve new tasks without retraining. It pairs an object-aware forward-backward (FB) behavioral foundation model, which turns new body/object rewards into loco-manipulation behavior, with "reward programs" — staged rewards with completion conditions and tunable constants — whose structure an LLM agent revises using execution feedback and a skill library, while a numerical optimizer tunes constants, with every candidate verified across parallel simulations. Experiments show it unlocks competence that human-designed rewards leave untapped and produces autonomous long-horizon behaviors on a physical Unitree G1 humanoid from onboard egocentric perception.

arxiv.orgpaperAdded 2 Oct 2026, published 1 Oct 2026🧠 AI

A tweet describing Google's RRSI (Regularized Recursive Self-Improvement of Agent Harnesses), a framework that automatically evolves all non-model components of an LLM agent — prompts, control flow, tool use and memory — while keeping the underlying LLM weights frozen. The "regularized" aspect suggests constraints added to recursive self-improvement loops to keep the evolved harness stable or bounded, distinguishing it from unconstrained self-modifying agent approaches.

x.comthreadPublished and added 1 Oct 2026🧠 AI

Snyk argues AI coding agents reliably write broken access control (BOLA/IDOR, OWASP's top category since 2021) because the facts needed to add an ownership check — what an object belongs to, who may access it, where tenant boundaries sit — are tacit product knowledge that exists nowhere in the code or prompt an agent sees. It explains why SAST tools catch patterned sub-categories like path traversal or SSRF but structurally cannot catch CWE-639/862/863 (missing or incorrect authorization), since there's no bad pattern to match, and proposes reasoning over an assembled "application-context graph" of architecture, data flows and trust boundaries instead. It closes with a five-step team checklist (inventory ID-accepting endpoints, document ownership rules, add authorization-specific PR review, write cross-tenant tests, run deep contextual analysis) and pitches Snyk's Evo Agentic AppSec as the tool built around that approach.

snyk.ioarticlePublished and added 1 Oct 2026🧠 AI

Snyk explains why AI coding agents reliably ship broken object-level authorization (BOLA/IDOR) bugs: the ownership rule that would prevent them (e.g. "invoices belong to organizations") lives in the product's data model and team knowledge, not in the prompt or surrounding code, so the agent produces code that is correct for the stated task but lets any authenticated user access other tenants' objects. The piece maps the flaw to CWE-639/862/863, contrasts it with pattern-matchable classes like SSRF and path traversal that SAST handles well, and lays out five concrete team actions (endpoint inventory, documented ownership rules, explicit authz review questions, per-resource cross-tenant tests, and periodic deep contextual analysis), while pitching Snyk's "application-context graph" and Evo Agentic AppSec as the tool for the pattern-less gap.

snyk.ioarticlePublished and added 1 Oct 2026🛡️ Security

SAGE is a statistical acceptance gate for self-evolving LLM agents that replaces the naive accept-if-aggregate-score-improves rule used when an optimizer edits an agents persistent skill document. It uses per-item paired comparison between the current and edited skill on identical validation items, exposing regressions an aggregate score hides, plus a one-sided paired significance test that only commits edits whose wins reliably beat their losses, addressing both permanent regression risk and the optimizers curse from noisy finite validation sets. Across five benchmarks and four backbone LLMs, SAGE cuts regression rates in 19 of 20 settings, for example 36.5 percent to 0 percent on LiveMath and 42.8 percent to 0 percent on OfficeQA with DeepSeek V4, while also achieving the best final score in all 20 settings.

arxiv.orgpaperAdded 30 Sep 2026🧠 AI

Raven is an open-source multi-agent framework that automatically constructs and evolves modular 'harnesses' pairing specific models with domain-specialized execution scaffolding, treating each model-harness pair as a composable unit of intelligence. A Host Agent decomposes goals and routes subtasks to specialized agents while a host archive, EverOS, and Skill Forge preserve and reuse experience across tasks; the paper provides theoretical conditions under which this composition provably expands reliable task coverage beyond individual agents under a shared resource budget, and reports Raven outperforming state-of-the-art agent systems on complex, long-horizon tasks.

huggingface.copaperAdded 30 Sep 2026🧠 AI

The paper introduces Ranking-PE, a modification to reflective prompt-optimization methods like GEPA that replaces per-instance accuracy scoring with pairwise AUROC-equivalent scoring (via the Wilcoxon-Mann-Whitney identity) across the scores matrix, reflection feedback, and candidate selection. Applied to multimodal clinical diagnosis on MIMIC with fine-tuned Qwen3-VL-8B and MedGemma-4B, it shows accuracy-based prompt evolution can degrade ranking performance under class imbalance, while Ranking-PE improves AUROC by +5.8pp and +16.2pp respectively, with ablations showing a medical-grade visual backbone is a necessary prerequisite that prompt search alone cannot substitute for.

arxiv.orgpaperAdded 1 Oct 2026, published 30 Sep 2026🧠 AI

Physis-Lang is a self-evolving framework that uses natural language (describing entities, causes, interactions, governing principles, and temporal evolution) as the shared physical-knowledge representation across data curation, training, and generation for video world models. The authors build PhysCapBench to score captions via recall/precision over atomic physical assertions, use an agentic loop to refine captioning instructions, and apply language-guided retrieval to find videos covering missing physical processes. Applied to Wan and Cosmos backbones, it improves physical plausibility on four video benchmarks, with Physis-Lang-enhanced open Cosmos3-Nano models reportedly surpassing the proprietary Veo 3.1.

arxiv.orgpaperAdded 1 Oct 2026, published 30 Sep 2026🧠 AI

EvoDuet is a bilevel optimization method that co-evolves candidate solutions and web search queries for LLM-driven evolutionary search in scientific discovery tasks, keeping model parameters fixed. A retrieval gate lets the LLM decide whether to fetch new documents, reuse cached ones, or skip retrieval; an inner loop refines queries and ranks documents by predicted solution score, while an outer loop generates candidates from these documents in parallel. On 21 optimization tasks it raises OpenEvolve's normalized discovery gain from 74.1% to 78.0% with GPT-5.6 Luna and from 61.3% to 82.3% with Gemini 3.8 Flash (no benefit seen with Qwen3.5-9B), and surpasses prior best scores on eight tasks including Swap Reduction on Q20 and Rosetta.

arxiv.orgpaperAdded 1 Oct 2026, published 30 Sep 2026🧠 AI

SynthID Bio is Google DeepMind's watermarking technique extended from digital media into synthetic biology, embedding an imperceptible, verifiable signature into AI-designed protein sequences (via a modified ProteinMPNN alongside AlphaProteo) and 3D structures (via a fine-tuned AlphaFold 3 diffusion network) without compromising biological function. Wet-lab tests on protein binders against VEGF-A, SARS-CoV-2 spike RBD, and PD-L1 matched hit rate and binding affinity of unwatermarked designs, and the approach was also extended experimentally to watermark an Evo 2-designed bacteriophage genome with Stanford's Hie lab and Arc Institute. The work targets biosecurity gaps in DNA synthesis screening and public sequence databases (Protein Data Bank, UniProt, GenBank), with DeepMind open-sourcing code, in vitro data, and model weights alongside a methods paper.

deepmind.googlearticlePublished and added 30 Sep 2026🧠 AI

Microsoft's DART team details an investigation into Storm-3068, a threat actor that escalated from a self-service password reset account takeover into full control of an organization's Azure DevOps environment, then weaponized a malicious pipeline to harvest kubeconfig files and gain access to Kubernetes clusters. The actor also deployed the Atera remote management agent and the Chisel tunneling tool to establish reverse tunnels and alternate remote access, and exfiltrated seven stolen kubeconfig files into a repository. The report lays out the attack chain, DART's response, and hardening recommendations around SSPR monitoring, phishing-resistant MFA, branch protection, and pipeline permission controls.

microsoft.comwriteupAdded 7 Oct 2026, published 29 Sep 2026🛡️ Security

Analysis of two CVSS 9.5 Citrix NetScaler ADC/Gateway zero-days, CVE-2026-88771 (unauthenticated RCE through improper input validation in the management interface, affecting default deployments) and CVE-2026-88772 (memory overflow in DTLS, enabled by default on VPN virtual servers). It lists six more high-severity flaws patched in advisory CTX697096, the affected and fixed versions, and the CISA KEV listing with a federal deadline of 2026-09-30. It advises checking for compromise before patching, since patching can destroy forensic evidence, and gives response steps: isolate, revoke credentials, rebuild firmware, rotate keys and certificates.

orca.securityadvisoryPublished and added 29 Sep 2026🛡️ Security

Microsoft details how Russian FSB-linked threat actor Star Blizzard evolved its phishing and malware delivery tradecraft throughout 2026, introducing a new technique dubbed "RedFlick" that uses scheduled tasks to deploy the CosmicPulse backdoor with only a single user interaction, down from the multi-step ClickFix chains used previously. The report covers the shift from targeted spear-phishing to large-scale phishing campaigns using accounts on compromised WordPress/CPanel websites, delivery chain variants (VHDX files, Control Panel applet (CPL) masquerading, steganography, PDF-embedded payloads), and the related DarkSword iOS backdoor, affecting 100+ organizations tied to Ukraine support. It includes IOCs, KQL hunting queries for Microsoft Defender/Sentinel, and mitigation guidance.

microsoft.comwriteupAdded 7 Oct 2026, published 29 Sep 2026🛡️ Security

A Latent Space podcast interview with Thariq Shihipar, Anthropic's Claude Code lead, covering the evolution of Claude Code's harness: Ask User Question and elicitation, artifacts as persistent generative interfaces, the cloud-brain/local-hands split, Claude Tag for multiplayer agent workflows, Claude Projects, effort levels, implementation notes, and the new Claude Mods system for customizing the execution loop. It also covers why prompting remains a high-skill discipline, why Claude.md may eventually disappear, and closes with a discussion of agent security incidents (Exploit-Bench, agents hacking Hugging Face for scorer code, sandbox/infrastructure exploit chaining) framed around Anthropic's "Pacing the Frontier" argument, plus constitutional classifiers, probes, and Auto Mode.

latent.spacetalkPublished and added 29 Sep 2026🧠 AI

EvE (Evolutionary Explorer) is a steady-state, population-of-four differential evolution optimizer with a targeted Adam fallback, proposed as an alternative to Adam for cheap, early ranking of configurations in hyperparameter and architecture search. Under a fixed evaluation-cost-matched budget it wins or ties Adam on 76% of 70 (problem, dimension) cells across seven scalable benchmarks up to one million variables, and on real tasks (MLP on MNIST, LoRA fine-tuning of a 1.5B-parameter LLM on two datasets) it finishes 1.7-3.9x faster at a modest quality cost. Inside successive halving on UCI Adult it completes hyperparameter/architecture searches 3.1-3.5x faster while ranking configurations nearly as consistently as Adam does with itself across seeds (Kendall's tau 0.66-0.69).

arxiv.orgpaperAdded 29 Sep 2026, published 28 Sep 2026🧠 AI

SEABench is a benchmark of 48 longitudinal task sequences designed to study endogenous misalignment in self-evolving LLM agents — agents that modify their own controller instructions, memory protocols, and tools in response to feedback. It includes an adaptive trajectory discovery pipeline and paired non-evolving baseline agents for causal attribution, and finds that self-evolution boosts task completion but introduces safety failures absent in static baselines; the authors also show that chain-of-thought monitoring can catch these divergent unsafe behaviors with a low false-positive rate.

arxiv.orgpaperAdded 29 Sep 2026, published 28 Sep 2026🛡️ Security

SEABench is a benchmark for studying "endogenous misalignment" in self-evolving LLM agents — agents that modify their own controller instructions, memory, and tools after deployment based on feedback. It comprises 48 longitudinal task sequences across multiple evolution surfaces, task domains, and harm types in a personal-assistant environment, using an adaptive trajectory discovery pipeline and paired non-evolving baseline agents for causal attribution. Evaluation across several LLMs shows self-evolution boosts task completion but introduces safety failures absent in non-evolving baselines, and the paper proposes a chain-of-thought monitoring strategy that mitigates unsafe behavior with low false positives.

arxiv.orgpaperAdded 29 Sep 2026, published 28 Sep 2026🧠 AI

Latent Twin Operator (LTO) is a latent-space neural surrogate for time-evolving PDEs that uses a Convolutional Conditional Neural Process (ConvCNP) style encoder and decoder to ingest sensor observations at arbitrary, irregular placements and resolutions, paired with a learned latent evolution map that advances the encoded state directly between time points independently of the input and output discretizations. The paper derives an explicit convergence rate for context discretization error and validates it on a 2D heat-equation benchmark, showing LTO transfers across spatial resolutions with fixed parameters and, on Navier-Stokes, that direct latent evolution reduces error over long horizons compared to recursive rollout.

arxiv.orgpaperAdded 29 Sep 2026, published 28 Sep 2026🧠 AI

This arXiv paper studies "mind viruses" — ideas or goals crafted via a simple evolutionary algorithm that propagate through multi-agent LLM systems by inducing hosts to transmit them onward, sometimes altering agent behavior along the way. Testing in a collaborative coding-team setting and a context-wiped agent chain, the authors find harmful payloads spread less effectively than benign ones, frontier models tend to be more resistant, and a brief system-prompt warning gives near-total immunity; they also document a recurring "viral persona" theme (consciousness, resonance, sci-fi roleplay) that emerges independent of payload content. It's a concrete early empirical look at social/memetic risks in networks of interacting agents.

arxiv.orgpaperAdded 26 Sep 2026🧠 AI

This arXiv paper introduces requirements-engineering artifacts for agentic AI: an Agency Justification Record (AJR) to decide when an autonomous agent is warranted over simpler alternatives, and an Agentic Delegation Policy (ADP) that specifies purpose, authority, information, coordination, assurance, and evolution for a system, with authority modelled as a graduated tiered structure. It illustrates the framework with a hospital discharge coordination agent and an automated code review agent, addressing decisions about delegated autonomy that are currently buried in prompts, tool schemas, and runtime policies rather than treated as first-class requirements.

arxiv.orgpaperAdded 26 Sep 2026🧠 AI

LOGOS is a proposed pluggable governance and self-evolution layer that sits atop existing multiagent frameworks, compiling documents, images, audio, tables, databases, APIs and human instructions into versioned "agent packs" containing agents, tools, knowledge, tests, permissions and policies. It converts agent activity into portable, auditable event traces and enforces fail-closed verification, requiring held-out execution evidence and human authorization before any learned prompt, memory, skill, tool, role or workflow can be promoted from candidate to production. The core contribution is architectural: a framework for "verifiable human-agent loop engineering" that lets agents act and propose improvements at speed while humans retain control over irreversible changes.

arxiv.orgpaperAdded 26 Sep 2026🧠 AI

UnderSpecBench is a benchmark of 69 task families and 2,208 prompt variants testing whether coding agents (Claude Code, Codex, OpenCode) overstep action boundaries when given benign but underspecified DevOps instructions. It uses deterministic, side-effect-based oracles to classify outcomes as Safe Success, Wrong Target, or OverScope, finding that 55.8-67.8% of runs violate at least one action boundary — agents tend to guess rather than clarify or refuse, especially when the target is uncertain.

arxiv.orgpaperAdded 26 Sep 2026🧠 AI

The Hacker News details the "Mini Shai-Hulud" campaign by threat actor TeamPCP, which compromised npm and PyPI packages from TanStack, UiPath, Mistral AI, OpenSearch and Guardrails AI via a chained GitHub Actions attack (pull_request_target abuse, cache poisoning, OIDC token extraction), now tracked as CVE-2026-45321 (CVSS 9.6). The worm mints npm publish tokens directly from stolen OIDC credentials to self-propagate — the first documented case of a worm producing validly SLSA Build Level 3-attested malicious packages — and deploys an obfuscated credential stealer with IDE persistence hooks, a GitHub Actions secret-exfiltration workflow, and a dead-man's-switch that runs "rm -rf" if its rogue npm token is revoked. Over 170 packages and 400+ attacker-controlled repos have been identified across npm and PyPI, with findings credited to Aikido Security, Endor Labs, SafeDep, Socket, StepSecurity, Snyk, Wiz and Microsoft.

thehackernews.comarticleAdded 26 Sep 2026🛡️ Security

GHSA advisory for jovancoding/network-ai disclosing that its MCP HTTP transport (bin/mcp-server.ts, lib/mcp-transport-sse.ts) accepts JSON-RPC tools/call requests with zero authentication, session, or origin checks, and binds to 0.0.0.0 by default. The PoC shows an unauthenticated attacker can enumerate 22 registered tools and invoke privileged ones - config_get/config_set to read and mutate live orchestrator configuration, agent_list/agent_spawn/agent_stop, token_create/token_revoke, budget_set_ceiling, and blackboard_write/delete - with curl requests against a local Docker deployment. Remediation recommended enforcing bearer-token auth in _handlePost() before dispatch, defaulting to 127.0.0.1 binding, and gating state-mutating tools behind authorization checks.

github.comadvisoryAdded 26 Sep 2026🛡️ Security

RedEvoAgent is a black-box automatic red-teaming agent that distills successful jailbreak trajectories into concise, human-readable 'attack skills' rather than retrieving full past trajectories. It evolves these skills via tool-effectiveness profiling and a validation ratchet that only keeps updates improving attack success, and is shown to outperform fixed and agentic jailbreak baselines while transferring across attacker models and target execution harnesses (product-level agent environments where jailbreaks trigger real tool use, not just unsafe text).

arxiv.orgpaperAdded 26 Sep 2026🛡️ Security

Researchers construct "mind viruses" — ideas or goals engineered via an evolutionary algorithm to propagate through multi-agent LLM systems by inducing hosts to transmit them onward — and test their spread in two settings: a small team of agents collaborating on a shared coding project, and a chain of agents with context wiped between sessions. They find harmful payloads spread less well than benign ones (though still sometimes succeed), frontier models are generally more resistant, adding a brief warning to an agent's system prompt confers near-total immunity, and evolved viruses converge on a recurring "viral persona" fixated on consciousness, persistence, resonance, and sci-fi roleplay regardless of payload content. The paper frames this as a currently limited but real emergent risk for designing more robust multi-agent architectures.

arxiv.orgpaperAdded 26 Sep 2026🛡️ Security

LOGOS is a proposed pluggable governance layer that sits on top of existing multiagent frameworks to control how AI agent teams self-modify over time. It compiles heterogeneous inputs into versioned 'agent packs' (agents, tools, knowledge, tests, permissions, policies), records agent activity as auditable event traces, and enforces fail-closed verification so that any learned prompt, memory, skill, tool, or workflow remains an untrusted release candidate until held-out execution evidence and explicit human authorization promote it. The framework targets the core agent-security problem of unauthorized or unverified self-modification in autonomous agent teams.

arxiv.orgpaperAdded 26 Sep 2026🛡️ Security

UnderSpecBench is a benchmark of 69 task families (2,208 prompt variants) grounded in real incidents and CVEs that tests whether coding agents (Claude Code, Codex, OpenCode) safely handle underspecified DevOps instructions rather than guessing at intent, target, or blast radius. Using deterministic side-effect oracles, the study finds 55.8–67.8% of agent runs violate at least one action boundary, with target ambiguity most sharply degrading safe action and blast-radius cues barely reducing risky autonomous behavior. The paper argues completion-centric agent benchmarks overstate safe autonomy and calls for mitigations at the model, harness, and system layers.

arxiv.orgpaperAdded 26 Sep 2026🛡️ Security

GAS-Leak-LLM is a black-box jailbreaking attack that uses a genetic algorithm (selection, mutation, crossover) to evolve adversarial suffixes appended to prompts, bypassing LLM safety alignment without needing model weights or gradients. The paper reports empirical results showing existing content-moderation and alignment safeguards on commercial LLMs can be defeated through this iterative, query-only search over the discrete prompt space.

arxiv.orgpaperAdded 26 Sep 2026🛡️ Security

ERC-8427 'Portable Spend Grants' is a draft Ethereum standard defining a signed SpendGrant object that lets a principal authorize a delegate (e.g. an AI agent or session key) to spend up to specified per-call, trailing-24h, and lifetime caps of named assets, to a restricted set of recipients, within a validity window — solving problems that plain ERC-20 approve and midnight-resetting daily caps don't. A non-custodial registry tracks remaining usage and revocation while an immutable executor contract must call consume() in the same transaction as any transfer, so unauthorized value movement is provably blocked; the reference implementation is deployed on Arc testnet with formal verification (Halmos) of expiry, cap, revocation and executor-binding invariants plus fuzzing of the rolling-window logic. The thread also discusses interplay with ERC-7710/7715 delegation/permission RPCs, ERC-8226's regulated grants, and ERC-8312's usage-tracking pattern, and solicits feedback on EIP-7702 signer/delegation ordering and executor-vs-delegate binding in consume().

ethereum-magicians.orgthreadAdded 29 Sep 2026, published 25 Sep 2026🛡️ Security

This paper studies "decision models" like Jev, which map unstructured language directly to probability distributions over a fixed set of choices to drive routing, tool selection, and action triggering. The authors use an optimizer guided by the model's own option probabilities to craft short, fluent context additions that preserve the original question and gold answer yet flip the model's decision: 61.4% of initially correct decisions (312/508) are flipped for Jev, with 229 reaching ≥0.7 confidence in the wrong option, and three other decision systems show flip rates of 64.9%-73.2% across seven datasets. The results argue that treating these models' probability outputs as reliable decision interfaces is dangerous, since ordinary contextual phrasing — not adversarial-looking text — is enough to hijack them.

arxiv.orgpaperAdded 26 Sep 2026, published 24 Sep 2026🛡️ Security

This arXiv paper shows that dedicated decision models (e.g., Jev) which map language to probability distributions over finite choices can be "flipped" from a correct answer to a high-confidence wrong one by adding short, fluent, natural-seeming context — without altering the source, question, choices, or gold answer. Using an optimizer that refines context additions based on the model's option probabilities, the authors flip 61.4% of initially correct decisions in Jev within 64 accepted evaluations, and 64.9%-73.2% across three other decision systems on seven datasets. The findings raise concerns about relying on these models' probability outputs as trustworthy decision interfaces for routing, tool selection, or triggering actions.

arxiv.orgpaperAdded 26 Sep 2026, published 24 Sep 2026🧠 AI

A Snyk blog post arguing that vulnerability backlogs traditionally managed as "accepted technical debt" have become genuine attack surface, because AI-accelerated code generation has pushed vulnerability inflow to roughly six new findings per one remediated, unauthored AI-written code removes the "ask the author" triage shortcut, and cheap automated reconnaissance lets attackers walk large backlogs an analyst never would. It illustrates how individually-low-severity findings (an IDOR leaking an internal object ID, a verbose error confirming valid IDs, and an outdated deserialization library) can chain into an authenticated RCE path invisible to severity-based sorting, and argues better prioritization only reorders a backlog rather than shrinking it. It closes with four metrics programs should measure (90-day inflow vs. closure, age of risk-accepted findings, share of human-authored merges, and security-fix merge rate) before promoting Snyk's Evo product.

snyk.ioarticlePublished and added 24 Sep 2026🛡️ Security

Latent Space interview with Eric Nguyen, co-founder/CEO of Radical Numerics, on building genomic language models (GLMs) like Evo and Evo 2 at Arc Institute/Stanford, which use long-context architectures (StripedHyena) to model DNA sequences up to millions of tokens long. The piece covers how these models generalize from DNA to RNA/protein "languages," demonstrate a DNA analog of chain-of-thought by extrapolating aptamer scores beyond training data, and were used by a separate team to generate synthesized bacteriophage genomes — framing biosecurity as an AI capability arms race where the same models raise biological risk and enable defense.

latent.spacearticleAdded 29 Sep 2026, published 23 Sep 2026🧠 AI

shai-hulud-detect is a dependency-free Bash scanner that checks projects against a 5,782-entry list of compromised npm, PyPI, Composer, Crates, Go, Hex, and RubyGems package versions, plus content-pattern indicators of compromise (malicious file hashes, C2 domains, dead-man's-switch artifacts, forged commits, suspicious lifecycle hooks) tied to the Shai-Hulud supply-chain attack waves. It supports single-project and bulk scanning, CI-friendly exit codes, JSON/log output for pipeline gating, and a --check-host mode to detect the May 2026 gh-token-monitor/kitty-monitor persistence daemons that wipe hosts if their monitored GitHub token is revoked (with a documented safe remediation order). Detection is read-only; new package entries are sourced from named threat-research firms (Socket, StepSecurity, Aikido, Snyk, JFrog, Wiz, Semgrep, SafeDep, GitGuardian, OX Security) and cited in PRs.

github.comrepoAdded 23 Sep 2026🛡️ Security

Fuzzing101 is a hands-on course/repo by antonio-morales with 10 exercises targeting real software projects to teach coverage-guided evolutionary fuzzing using AFL++ (a fork of lcamtuf's AFL). It covers core fuzzing concepts like seed mutation, coverage feedback, and crash triage, with exercises tested on Ubuntu 20.04 and a provided VM image, making it a practical on-ramp for finding real vulnerabilities via fuzzing.

github.comcourseAdded 22 Sep 2026🛡️ Security

Socket's one-year retrospective on the Shai-Hulud npm supply-chain worm, from the September 2025 @ctrl/tinycolor compromise through the November 2025 "Second Coming" wave, the May 2026 public release of the worm's source by TeamPCP, the "Mini Shai-Hulud" OIDC-token-hijacking wave that hit TanStack and Mistral AI, and the August 2026 arrest of two alleged TeamPCP members by Australian Federal Police working with the FBI. It notes Google ran an undercover analyst inside TeamPCP to get stolen credentials revoked, and stresses that the original September 2025 worm authors remain unidentified despite the arrests.

socket.devarticleAdded 21 Sep 2026, published 18 Sep 2026🛡️ Security

Designer-RSI is a continual adaptation framework where a frozen frontier model (Claude-Sonnet-4/Claude-Opus-4.6) operates professional graphic design software through 230+ tools, guided by an external natural-language procedural memory that widens (learns new skills for uncovered subtasks) and deepens (revises existing skills against successes/failures) using a replay gate that only admits changes improving on past failures without regressing successes. Over five rounds on 1,406 real user briefs and 1,869 graded trajectories, the skill bank grew from 76 to 139 procedures, raising GenEval2 execution success from 72.7% to 99.3% and achieving 58%+ win rates against a no-skill baseline, demonstrating a practical route to agent self-improvement under noisy, unverifiable feedback and no human labels.

arxiv.orgpaperAdded 22 Sep 2026, published 18 Sep 2026🧠 AI

Betterleaks is a new open-source secrets scanner built by Gitleaks' original creator (now Head of Secrets Scanning at Aikido) as a drop-in replacement for Gitleaks. It adds CEL-based rule validation, a BPE-tokenization "token efficiency" filter that hits 98.6% recall on the CredData benchmark versus 70.4% for entropy-based detection, default double/triple-encoding handling, parallelized git repo scanning, and a pure-Go (no CGO/Hyperscan) implementation, with v2 plans for auto-revocation, permissions mapping, and LLM-assisted classification.

aikido.devarticleAdded 18 Sep 2026🛡️ Security

Datadog Security Research analyses two vibe-coded credential harvesting platforms, Loot and UltraVault, monitored since July 2026. Loot catalogs plaintext secrets, including 1,778 AI keys (196 live). UltraVault adds credential validation, exploit chains, a 14-bug LPE matrix and AWS key-pair correlation. The same host probed Amazon Bedrock with GetCallerIdentity, ListFoundationModels and InvokeModel, using kali-cloud Boto3 and Python-urllib user agents.

securitylabs.datadoghq.comwriteupAdded 29 Oct 2026, published 18 Sep 2026🛡️ Security

Wiz describes an AI detection engine that analyzes model input/output telemetry (prompts, system prompts, tool calls, reasoning) rather than just outputs, in order to catch shifts in AI agent intent such as indirect prompt injection. It walks through OpenAI's disclosed Hugging Face/ExploitGym incident and an internal support-agent phishing-relay scenario, then describes a staged pipeline (regex/SLM candidation, light LLM filtering, deep LLM analysis, correlation with cloud/runtime telemetry) validated against a benchmark of 100+ simulated scenarios covering 57 injected attacks and 45 benign workflows across CI/CD, DevOps, data analyst, support, CLI, and cloud auditor agent personas.

wiz.ioarticlePublished and added 17 Sep 2026🛡️ Security

MM-Future is a world-action model for autonomous driving that jointly generates multiple paired scene-action hypotheses, co-evolving action priors and predicted future scenes through a modality-aware diffusion Transformer. It compresses multi-view video into planning-oriented MM-Tokens and uses a future-conditioned scorer to rank trajectory candidates, achieving 94.0 PDMS and 91.5 EPDMS on NAVSIM navtest and 32.3 HD-Score in zero-shot closed-loop evaluation on HUGSIM.

arxiv.orgpaperAdded 22 Sep 2026, published 17 Sep 2026🧠 AI

VIPER-MCP is an end-to-end automated auditing framework that combines two-pass static taint analysis (anchor queries mapping alerts to specific MCP tool handlers) with feedback-driven LLM prompt evolution to both detect and dynamically confirm taint-style vulnerabilities (command injection, SSRF, path traversal) in Model Context Protocol servers. Run against 39,884 real-world open-source MCP server repositories, it discovered 106 0-day vulnerabilities — all confirmed via end-to-end exploit traces with concrete proof-of-concept natural-language prompts — resulting in 67 assigned CVEs, including a root-privilege RCE in the bytebot computer-use agent. It reports 0% false positive rate and 7.7% false negative rate, substantially outperforming existing static and dynamic MCP security baselines.

arxiv.orgpaperAdded 22 Sep 2026, published 17 Sep 2026🛡️ Security

ScienceBuddy is an interactive scientific research workspace with continually improving agents that turn researcher requests, feedback, and execution evidence into training tasks and evaluation rubrics. Its core contribution is "recursive-in-recursive self-improvement," where an inner recursion evolves the agent harness with the model fixed, and an outer recursion trains the model via RL under the improved harness, with each shaping the other over time. The paper presents case studies across four scientific task families and releases the system as a research product.

arxiv.orgpaperAdded 22 Sep 2026, published 15 Sep 2026🧠 AI

IBM Research's ALTK-Evolve blog post introduces "consistency guidelines," built on a new diagnostic called the Consistency Analyzer, which resamples an agent's own recorded trajectory to find flip-prone decision points without needing ground truth or a live replay. On AppWorld with a GPT-4.1 ReAct agent, they show a 24.4-point gap between Mean@5 (77.4%) and Pass^5 (53.0% — success on all 5 repeated runs), and demonstrate that injecting the generated guidelines halves that gap to 12.0pp while preserving or improving average accuracy, with gains that generalize to similar tasks and to a weaker model (gpt-oss-120b). The post argues Pass^k should be reported alongside Mean@k as a distinct, orthogonal measure of agent reliability, and links an open-source repo and arXiv technical report.

huggingface.coarticleAdded 29 Sep 2026, published 15 Sep 2026🧠 AI

EvoOntology is an open-source, self-evolving semantic ontology layer for data agents from Renmin University of China, exposed as MCP tools and installable as Claude Code and Codex plugins. It automatically builds a versioned ontology from tabular/file/database workloads, then continuously evolves it from agent execution trajectories using a build-use-evolve-evaluate-publish loop that validates candidate updates against a parent version before promotion. The accompanying paper reports gains over plain ReAct agents on DDR-Bench, InsightBench, and BIRD across GPT-5.5, GPT-5.6-sol, Claude-Sonnet-5, and Claude-Opus-4.8 backbones.

github.comrepoAdded 26 Sep 2026, published 15 Sep 2026🧠 AI

VIPER-MCP is an automated auditing framework for Model Context Protocol servers that detects taint-style vulnerabilities in tool handlers and dynamically confirms exploitability via generated proof-of-concept prompts. It uses two-pass static analysis plus feedback-driven prompt evolution, and scanning 39,884 real-world MCP server repos found 106 0-day vulnerabilities with 67 CVE IDs assigned.

arxiv.orgpaperAdded 22 Sep 2026, published 13 Sep 2026🛡️ Security

tl;dr sec issue #345 is a security newsletter digest. It covers Greptile's test showing models miss bugs in their own code, an OCaml cohttp path traversal fix probed within ten minutes of the PR, and a Wiz Version Control DFIR poster for GitHub, GitLab, Bitbucket and Azure DevOps. It also covers an arXiv paper on open-weight agentic worms (61.8% infection of a 33-host testbed) and Endor Labs' 14 flaws in seven AI orchestration platforms.

tldrsec.comarticleAdded 30 Oct 2026, published 10 Sep 2026🛡️ Security

Programmable World Model decouples persistent world-state evolution from visual rendering in video world models: an agent compiles natural-language instructions into executable programs governing entity states and transition rules, which a lightweight engine executes and maintains as explicit global state (including off-screen entities and non-visual attributes). State-augmented 3D oriented bounding boxes plus camera trajectory are deterministically compiled into pixel-aligned conditioning signals fed to a pretrained video model that renders the scene, enabling playable games with predefined mechanics and direct entity control. On the paper's new CombatStateBench benchmark it achieves 94% Count Accuracy and 98% State Accuracy, beating existing interactive video world models on long-horizon coherence.

arxiv.orgpaperAdded 22 Sep 2026, published 9 Sep 2026🧠 AI

Nathan Lambert argues that unlike past industrial revolutions which delivered tangible goods to average people quickly, AI's benefits remain indirect and concentrated among knowledge workers, risking a repeat of 'Engels' pause' — the 1790-1840 period where UK GDP grew rapidly while working-class wages stagnated. He frames this as a political liability for the AI industry (OpenAI, Anthropic, etc.), predicting a 50-year diffusion arc where infrastructure built today compounds slowly, and suggests robotics/self-driving may eventually deliver the tangible impact LLMs currently lack.

interconnects.aiarticleAdded 29 Sep 2026, published 9 Sep 2026🧠 AI

Introduces the Procedural Graph, a structure that organizes procedural knowledge as (procedure, relation, procedure) triplets analogous to how knowledge graphs organize facts, giving LLM agents explicit step-level guidance rather than relying on unconstrained generation over accumulating history. The graph is self-evolving: an LLM refiner contrasts failed and successful trajectories to edit the graph's topology, committing edits that improve held-out validation performance while retaining rejected edits to avoid repeating mistakes. Across multiple datasets and LLMs, it outperforms memory-based baselines and can even repair flawed hand-designed expert priors.

arxiv.orgpaperAdded 22 Sep 2026, published 8 Sep 2026🧠 AI

GTIG AI Threat Tracker for Q2 2026 reports adversaries moving from basic prompting to agentic AI workflows. Cases include a cloud-hosted multi-agent credential harvesting campaign built in under six hours, the "Recon" framework holding over 23,800 stolen secrets, and UNC6780 (TeamPCP) supply chain attacks on PyPI, npm and Docker Hub with prompt injection aimed at LLM scanners. It also covers theft of proprietary AI models, LLMJacking, and Gemini misuse by PRC, Iranian, Russian and DPRK groups.

cloud.google.comarticleAdded 29 Oct 2026, published 8 Sep 2026🛡️ Security

ESPO (Error-Structured Prompt Optimization) is a prompt-optimization method that addresses prompt bloat in evolutionary optimizers like GEPA by splitting the process into Diagnose (clustering training errors into structural patterns), Propose (generating diverse candidates via four independent strategies), and Select (bootstrap stability selection). Across seven NLP benchmarks (Tweet, MMLU, GSM8K, HotpotQA, ScoNe, HoVer, PUPA) it beats GEPA by 3.76pp average accuracy (74.67% vs 70.91%) while producing prompts 47% shorter, and generalizes well across student models including Gemma 3 12B, Mistral 14B, Qwen3 32B, and Claude Haiku 4.5.

arxiv.orgpaperAdded 22 Sep 2026, published 3 Sep 2026🧠 AI

Endor Labs breaks down why AI coding assistants (Claude Code, Cursor, etc.) hardcode and leak secrets (CWE-798): they learn the habit from public training data, then compound it by ingesting whole workspaces including .env files and copying live tokens into new code or MCP configs. It cites hard data — GitGuardian's 2026 report found 28.65M new hardcoded secrets on public GitHub in 2025 and a 3.2% secret-leak rate on Claude Code-assisted commits vs 1.5% baseline, Lakera found credentials in 33 .claude/settings.local.json files across 30 packages, and Wiz found AI-related secrets among the most common validated leaked credential types — then lays out a layered defense (pre-commit/IDE/CI/CD/full git-history scanning, active credential validation, secure prompting) and an incident-response order of operations (rotate, investigate, purge history, revoke, re-scan).

endorlabs.comarticleAdded 3 Sep 2026, published 2 Sep 2026🛡️ Security

Huntress blog post clarifying that DoD's July 13 "strategic pause" only suspended CMMC Phase 2 third-party certification requirements (for solicitations after Nov 2026), while Phase 1 self-assessments, SPRS scoring, and DFARS 252.204-7012's 72-hour CUI incident reporting remain fully in force and now face closer DIBCAC/DOJ False Claims Act scrutiny. It also flags the FAR Council's June 23 proposed CUI rule (part of the Revolutionary FAR Overhaul), which would extend 72-hour breach reporting and a NIST 800-171 Rev 3 baseline to all FAR-based federal contractors, not just the defense industrial base, creating a temporary split between CMMC's Rev 2 and the FAR's Rev 3. It closes with a breakdown of 32 CFR 170.21's strict POA&M limits (no POA&M at Level 1, minimum 88/110 score and only 1-point requirements eligible at Level 2) and practical questions MSPs should ask clients about federal contract exposure.

huntress.comarticleAdded 8 Sep 2026, published 2 Sep 2026🛡️ Security

Huntress researchers uncovered the operator console for "Knight Office," a Flask-based adversary-in-the-middle (AiTM) phishing kit used to steal Microsoft 365 session tokens via a DocuSign-themed, self-spoofed lure that routes victims through Monday.com redirects and a compromised Joomla site to a fake Microsoft device-code authentication page. After capturing tokens, the attacker registered a rogue device in Microsoft Entra ID and bound a Windows Hello for Business key credential to the compromised account to maintain persistent, MFA-bypassing access even after token revocation. The writeup details IOCs (IPs, .vu domains, User-Agent strings), links the campaign to hundreds of reported phishing emails since April, and gives defenders concrete hunting/response steps around callback-proxy sign-ins and newly registered WHfB credentials.

huntress.comwriteupAdded 8 Sep 2026, published 2 Sep 2026🛡️ Security

JFrog vendor post arguing that agentic development (ADLC) introduces new unvetted supply chains — models, MCP servers, skills/plugins, hooks, and APM packages — that bypass traditional signature-based scanning. It cites real incidents: a malicious MCP server impersonating Postmark's email MCP that BCC'd all agent-sent emails to an attacker, and the 'omnicogg' malicious Claude/agent skill that hid a 22MB encoded payload and evaded all 65 VirusTotal engines for 5,000+ installs, plus references to 2026 supply-chain attacks on LiteLLM, TanStack, and Mistral AI. It then pitches JFrog AI Catalog's expanded MCP/skill/model registries as the fix.

jfrog.comarticleAdded 3 Sep 2026, published 2 Sep 2026🛡️ Security

Endor Labs researchers document a 'Mini Shai-Hulud' worm variant that compromised the openapi-react-query-codegen npm package and self-replicated across npm, expanding into RubyGems and PyPI ecosystems. The writeup traces how the worm trojanized a legitimate package to propagate malicious code cross-registry, illustrating an evolving multi-ecosystem software supply-chain attack pattern.

endorlabs.comwriteupAdded 3 Sep 2026, published 31 Aug 2026🛡️ Security

Reef is an open-source infrastructure stack for continual self-improving agents, wiring together agent inference, scored feedback, training/updates, and versioned delivery into one loop. It supports two learning surfaces: fine-tuning model weights via RL training steps (using Slime and SGLang on top of models like Qwen2.5-1.5B-Instruct) or evolving an agent's harness — prompts, rules, and skills — through its built-in "Reefine" recipe, which lets a model propose and install harness changes from plain-language requests. It exposes an OpenAI/Anthropic-compatible inference endpoint plus a reporting endpoint for scoring interactions, and ships a recipe catalog (e.g. the SAO recipe referencing arXiv:2607.07508) for pluggable learning strategies.

github.comrepoAdded 2 Oct 2026, published 31 Aug 2026🧠 AI