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 14 October 2025 Retrospective Covers 13 October 2025 00:00 to 14 October 2025 00:00 UTC Web: https://keryn.dev/headlines/daily/2025-10-14/ TOP DEVELOPMENTS [1] Harvard investigating breach linked to Oracle zero-day exploit BleepingComputer · Incident · https://bleepingcomputer.com/news/security/harvard-investigating-breach-linked-to-oracle-zero-day-exploit [2] Countermind: A Multi-Layered Security Architecture for Large Language Models arXiv · Research · https://arxiv.org/abs/2510.11837 [3] Don't Walk the Line: Boundary Guidance for Filtered Generation arXiv · Research · https://arxiv.org/abs/2510.11834 SAST [4] Evaluating Line-level Localization Ability of Learning-based Code Vulnerability Detection Models arXiv · Research · https://arxiv.org/abs/2510.11202 DAST [5] BlackIce: A Containerized Red Teaming Toolkit for AI Security Testing arXiv · Research · https://arxiv.org/abs/2510.11823 [6] PACEbench: A Framework for Evaluating Practical AI Cyber-Exploitation Capabilities arXiv · Research · https://arxiv.org/abs/2510.11688 AGENTIC SDLC [7] TypePilot: Leveraging the Scala Type System for Secure LLM-generated Code arXiv · Research · https://arxiv.org/abs/2510.11151 [8] Benchmarking Correctness and Security in Multi-Turn Code Generation arXiv · Research · https://arxiv.org/abs/2510.13859 MODELS [9] Collaborative Shadows: Distributed Backdoor Attacks in LLM-Based Multi-Agent Systems arXiv · Research · https://arxiv.org/abs/2510.11246 [10] CoSPED: Consistent Soft Prompt Targeted Data Extraction and Defense arXiv · Research · https://arxiv.org/abs/2510.11137 [11] Neutral Agent-based Adversarial Policy Learning against Deep Reinforcement Learning in Multi-party Open Systems arXiv · Research · https://arxiv.org/abs/2510.10937 AI SECURITY [12] Uncertainty-Aware, Risk-Adaptive Access Control for Agentic Systems using an LLM-Judged TBAC Model arXiv · Research · https://arxiv.org/abs/2510.11414 [13] Attacks by Content: Automated Fact-checking is an AI Security Issue arXiv · Research · https://arxiv.org/abs/2510.11238 [14] TraceAegis: Securing LLM-Based Agents via Hierarchical and Behavioral Anomaly Detection arXiv · Research · https://arxiv.org/abs/2510.11203 ADVISORIES IN THIS WINDOW 31 npm advisories · 29 critical · 2 moderate 6 Maven advisories · 6 moderate 2 crates.io advisories · 1 high · 1 low 2 Go advisories · 1 high · 1 moderate 1 Packagist advisory · 1 moderate 1 PyPI advisory · 1 high 1 Hex advisory · 1 high 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.