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 Daily, Tuesday 29 September 2026 Retrospective Covers 28 September 2026 00:00 to 29 September 2026 00:00 UTC Web: https://keryn.dev/headlines/daily/2026-09-29/ TOP DEVELOPMENTS [1] 80,000+ Organizations Had AI Logins Stolen: From Shadow AI to LLMjacking BleepingComputer · Incident · https://bleepingcomputer.com/news/security/80-000-plus-organizations-had-ai-logins-stolen-from-shadow-ai-to-llmjacking [2] Apple Patches CoreGraphics Flaw Possibly Exploited in Targeted Attacks The Hacker News · Advisory · https://thehackernews.com/2026/09/apple-patches-coregraphics-flaw.html [3] NeedyMantis: Unpacking a post-compromise malware family used in targeted operations Microsoft · Incident · https://microsoft.com/en-us/security/blog/2026/09/28/needymantis-unpacking-a-post-compromise-malware-family-used-in-targeted-operations SAST [4] Trajectory-Level Security Debt in LLM Coding Agents arXiv · Research · https://arxiv.org/abs/2609.35199 DAST [5] TraceLib: System-Call Bitmap Feedback Mechanism for Language-Agnostic Web Fuzzing arXiv · Research · https://arxiv.org/abs/2609.34778 AGENTIC SDLC [6] Assay: Claims That Decay With the Code. Content-Addressed Evidence Graphs for Accountable AI-Assisted Software Delivery arXiv · Research · https://arxiv.org/abs/2609.36170 [7] The Compiler May Read It, the Agent May Not: Keeping Part of a Research Code Away from a Coding Agent arXiv · Research · https://arxiv.org/abs/2609.35557 [8] Continuous Assurance of Agentic Security Auditors for Software Delivery Decision Gates arXiv · Research · https://arxiv.org/abs/2609.35266 REMEDIATION [9] ReproBench: Benchmarking LLM Agents on Reproducing Vulnerability From Scratch arXiv · Research · https://arxiv.org/abs/2609.34450 [10] Maintaining Benchmarks Against Increasingly Capable Agents: Detection and Remediation of Unearned Passes arXiv · Research · https://arxiv.org/abs/2609.34262 [11] Athena's disclosures begin Chainguard · Analysis · https://chainguard.dev/unchained/athenas-disclosures-begin MODELS [12] Jailbreaks for Black-Box Uncertainty Quantification in Large Reasoning Models arXiv · Research · https://arxiv.org/abs/2609.35350 [13] Narrow Multimodal Fine-Tuning Can Induce Emergent Misalignment arXiv · Research · https://arxiv.org/abs/2609.35291 [14] Physics-Attested Federated Learning: Securing Collaborative Anomaly Detection in Critical Water Infrastructure arXiv · Research · https://arxiv.org/abs/2609.34804 AI SECURITY [15] Render Before Reading: Visual Rendering as a Prompt Injection Defense arXiv · Research · https://arxiv.org/abs/2609.36121 [16] Same Bytes, Different Authority: Reserved-Token Representations in Chat-Template Prompt Injection arXiv · Research · https://arxiv.org/abs/2609.35932 [17] CoSec: Benchmarking Agent Security in Communities arXiv · Research · https://arxiv.org/abs/2609.34790 ADVISORIES IN THIS WINDOW 56 npm advisories · 43 critical · 5 high · 8 moderate 3 Maven advisories · 1 high · 2 moderate 1 PyPI advisory · 1 critical 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.