PREVIEW. This is how it will look in an inbox; it was not sent to you. Each real email carries its own unsubscribe link and Keryn's postal address. KERYN HEADLINES Weekly, week 3 of 2026 Retrospective Covers 6 January 2026 12:00 to 13 January 2026 12:00 UTC Web: https://keryn.dev/headlines/weekly/2026-W03/ TOP DEVELOPMENTS [1] CISA orders feds to patch Gogs RCE flaw exploited in zero-day attacks BleepingComputer · Advisory · https://bleepingcomputer.com/news/security/cisa-orders-feds-to-patch-gogs-rce-flaw-exploited-in-zero-day-attacks [2] Cisco warns of Identity Service Engine flaw with exploit code BleepingComputer · Research · https://bleepingcomputer.com/news/security/cisco-warns-of-identity-service-engine-flaw-with-exploit-code [3] Veeam issues patch to close critical remote code execution flaw CyberScoop · Advisory · https://cyberscoop.com/veeam-backup-replication-security-flaw-remote-code-execution-fix [4] n8n Supply Chain Attack Abuses Community Nodes to Steal OAuth Tokens The Hacker News · Incident · https://thehackernews.com/2026/01/n8n-supply-chain-attack-abuses.html [5] ServiceNow Patches Critical AI Platform Flaw Allowing Unauthenticated User Impersonation The Hacker News · Advisory · https://thehackernews.com/2026/01/servicenow-patches-critical-ai-platform.html SAST [6] Towards Compositional Generalization in LLMs for Smart Contract Security: A Case Study on Reentrancy Vulnerabilities arXiv · Research · https://arxiv.org/abs/2601.06914 [7] HogVul: Black-box Adversarial Code Generation Framework Against LM-based Vulnerability Detectors arXiv · Research · https://arxiv.org/abs/2601.05587 [8] AutoVulnPHP: LLM-Powered Two-Stage PHP Vulnerability Detection and Automated Localization arXiv · Research · https://arxiv.org/abs/2601.06177 SCA [9] CHASE: LLM Agents for Dissecting Malicious PyPI Packages arXiv · Research · https://arxiv.org/abs/2601.06838 [10] A Large Scale Empirical Analysis on the Adherence Gap between Standards and Tools in SBOM arXiv · Research · https://arxiv.org/abs/2601.05622 [11] LIDL: LLM Integration Defect Localization via Knowledge Graph-Enhanced Multi-Agent Analysis arXiv · Research · https://arxiv.org/abs/2601.05539 [12] Researchers Uncover NodeCordRAT Hidden in npm Bitcoin-Themed Packages The Hacker News · Research · https://thehackernews.com/2026/01/researchers-uncover-nodecordrat-hidden.html DAST [13] PenForge: On-the-Fly Expert Agent Construction for Automated Penetration Testing arXiv · Research · https://arxiv.org/abs/2601.06910 [14] Cybersecurity AI: A Game-Theoretic AI for Guiding Attack and Defense arXiv · Research · https://arxiv.org/abs/2601.05887 AGENTIC SDLC [15] Datadog uses Codex for system-level code review OpenAI News · Analysis · https://openai.com/index/datadog AI-BOM [16] AIBoMGen: Generating an AI Bill of Materials for Secure, Transparent, and Compliant Model Training arXiv · Research · https://arxiv.org/abs/2601.05703 [17] The AI Inventory Gap: Why Your Organization Has No Idea What AI Assets Are Part of Your Software Supply Chain Checkmarx · Analysis · https://checkmarx.com/blog/ai-llm-tools-in-application-security/the-ai-inventory-gap-why-your-organization-has-no-idea-what-ai-assets-are-part-of-your-software-supply-chain REMEDIATION [18] Trend Micro warns of critical Apex Central RCE vulnerability BleepingComputer · Analysis · https://bleepingcomputer.com/news/security/trend-micro-fixes-critical-rce-flaw-in-apex-central-console [19] New Veeam vulnerabilities expose backup servers to RCE attacks BleepingComputer · Analysis · https://bleepingcomputer.com/news/security/new-veeam-vulnerabilities-expose-backup-servers-to-rce-attacks [20] SecureCAI: Injection-Resilient LLM Assistants for Cybersecurity Operations arXiv · Research · https://arxiv.org/abs/2601.07835 [21] Signal in the Noise: An Industry-Wide Perspective on the State of VEX OpenSSF · Analysis · https://openssf.org/blog/2026/01/08/signal-in-the-noise-an-industry-wide-perspective-on-the-state-of-vex TRUST [22] Evaluating Implicit Regulatory Compliance in LLM Tool Invocation via Logic-Guided Synthesis arXiv · Research · https://arxiv.org/abs/2601.08196 [23] R-LAM: Reproducibility-Constrained Large Action Models for Scientific Workflow Automation arXiv · Research · https://arxiv.org/abs/2601.09749 [24] Brokerage in the Black Box: Swing States, Strategic Ambiguity, and the Global Politics of AI Governance arXiv · Research · https://arxiv.org/abs/2601.06412 [25] Bathtubs, Boundaries, and Sandboxes: AI Regulatory Learning under Legal Uncertainty arXiv · Research · https://arxiv.org/abs/2601.04094 MODELS [26] ForgetMark: Stealthy Fingerprint Embedding via Targeted Unlearning in Language Models arXiv · Research · https://arxiv.org/abs/2601.08189 [27] SAGE: Tool-Augmented LLM Task Solving Strategies in Scalable Multi-Agent Environments arXiv · Research · https://arxiv.org/abs/2601.09750 [28] When Bots Take the Bait: Exposing and Mitigating the Emerging Social Engineering Attack in Web Automation Agent arXiv · Research · https://arxiv.org/abs/2601.07263 [29] MacPrompt: Maraconic-guided Jailbreak against Text-to-Image Models arXiv · Research · https://arxiv.org/abs/2601.07141 AI SECURITY [30] WebTrap Park: An Automated Platform for Systematic Security Evaluation of Web Agents arXiv · Research · https://arxiv.org/abs/2601.08406 [31] Superhuman AI Exfiltrates Emails Simon Willison · Analysis · https://simonwillison.net/2026/Jan/12/superhuman-ai-exfiltrates-emails [32] Reasoning Models Will Sometimes Lie About Their Reasoning arXiv · Research · https://arxiv.org/abs/2601.07663 [33] MCP-ITP: An Automated Framework for Implicit Tool Poisoning in MCP arXiv · Research · https://arxiv.org/abs/2601.07395 ADVISORIES IN THIS WINDOW 178 npm advisories · 155 critical · 12 high · 6 moderate · 5 low 29 PyPI advisories · 3 critical · 11 high · 10 moderate · 5 low 12 crates.io advisories · 5 high · 3 moderate · 4 low 10 Maven advisories · 3 critical · 2 high · 5 moderate 9 Go advisories · 3 critical · 6 moderate 5 Packagist advisories · 5 moderate 3 Multi-ecosystem advisories · 2 high · 1 low 2 Swift advisories · 1 high · 1 low 2 RubyGems advisories · 1 high · 1 moderate 1 NuGet advisory · 1 low All advisories: https://keryn.dev/headlines/advisories/ Titles, links and short excerpts only; every item links to its source. Preview: the unsubscribe link and Keryn's postal address are added to each email when it is sent.