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, Monday 15 December 2025 · Retrospective · Covers 14 December 2025 00:00 to 15 December 2025 00:00 UTC

Read on the web

Top developments

Auto-Tuning Safety Guardrails for Black-Box Large Language Models

arXiv · Research · Models

The Laminar Flow Hypothesis: Detecting Jailbreaks via Semantic Turbulence in Large Language Models

arXiv · Research · Models

From Linear Risk to Emergent Harm: Complexity as the Missing Core of AI Governance

arXiv · Research · Trust

SAST

SHERLOCK: A Deep Learning Approach To Detect Software Vulnerabilities

arXiv · Research · SAST

SCA

Unveiling Malicious Logic: Towards a Statement-Level Taxonomy and Dataset for Securing Python Packages

arXiv · Research · SCA

Cargo Sherlock: An SMT-Based Checker for Software Trust Costs

arXiv · Research · SCA

Models

One Leak Away: How Pretrained Model Exposure Amplifies Jailbreak Risks in Finetuned LLMs

arXiv · Research · AI Security, Models

AI Security

ceLLMate: Sandboxing Browser AI Agents

arXiv · Research · AI Security

Detecting Prompt Injection Attacks Against Application Using Classifiers

arXiv · Research · AI Security

Advisories in this window

  • 1 Maven advisory · 1 moderate
  • 1 crates.io advisory · 1 moderate

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.