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

Read on the web

Top developments

Harvard investigating breach linked to Oracle zero-day exploit

BleepingComputer · Incident

Countermind: A Multi-Layered Security Architecture for Large Language Models

arXiv · Research · AI Security, Models, Trust

Don't Walk the Line: Boundary Guidance for Filtered Generation

arXiv · Research · Models

SAST

Evaluating Line-level Localization Ability of Learning-based Code Vulnerability Detection Models

arXiv · Research · SAST

DAST

BlackIce: A Containerized Red Teaming Toolkit for AI Security Testing

arXiv · Research · DAST

PACEbench: A Framework for Evaluating Practical AI Cyber-Exploitation Capabilities

arXiv · Research · DAST

Agentic SDLC

TypePilot: Leveraging the Scala Type System for Secure LLM-generated Code

arXiv · Research · Agentic SDLC

Benchmarking Correctness and Security in Multi-Turn Code Generation

arXiv · Research · Agentic SDLC

Models

Collaborative Shadows: Distributed Backdoor Attacks in LLM-Based Multi-Agent Systems

arXiv · Research · Models

CoSPED: Consistent Soft Prompt Targeted Data Extraction and Defense

arXiv · Research · Models

Neutral Agent-based Adversarial Policy Learning against Deep Reinforcement Learning in Multi-party Open Systems

arXiv · Research · Models

AI Security

Uncertainty-Aware, Risk-Adaptive Access Control for Agentic Systems using an LLM-Judged TBAC Model

arXiv · Research · AI Security

Attacks by Content: Automated Fact-checking is an AI Security Issue

arXiv · Research · AI Security

TraceAegis: Securing LLM-Based Agents via Hierarchical and Behavioral Anomaly Detection

arXiv · Research · AI Security

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

Titles, links and short excerpts only; every item links to its source. Selection is automatic; the editor's note is AI-drafted and checked as described. Preview: the unsubscribe link and Keryn's postal address are added to each email when it is sent.