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 2 of 2025 · Retrospective · Covers 31 December 2024 12:00 to 7 January 2025 12:00 UTC

Read on the web

Top developments

Malicious npm packages target Ethereum developers' private keys

BleepingComputer · Analysis · SCA

Bad Likert Judge: A Novel Multi-Turn Technique to Jailbreak LLMs by Misusing Their Evaluation Capability

Palo Alto Networks · Analysis · Models

Malicious Obfuscated NPM Package Disguised as an Ethereum Tool Deploys Quasar RAT

The Hacker News · Research · SCA

SAST

CommitShield: Tracking Vulnerability Introduction and Fix in Version Control Systems

arXiv · Research · Remediation, SAST

LLM4CVE: Enabling Iterative Automated Vulnerability Repair with Large Language Models

arXiv · Research · Agentic SDLC, Remediation, SAST

SCA

Leveraging Large Language Models and Machine Learning for Smart Contract Vulnerability Detection

arXiv · Research · SCA

Agentic SDLC

Microsoft 365 Copilot Generated Images Accessible Without Authentication -- Fixed!

Embrace The Red · Analysis · Agentic SDLC

Remediation

Layer-Level Self-Exposure and Patch: Affirmative Token Mitigation for Jailbreak Attack Defense

arXiv · Research · Models, Remediation

Predicting IoT Device Vulnerability Fix Times with Survival and Failure Time Models

arXiv · Research · Remediation

LDAPNightmare PoC Exploit Crashes LSASS and Reboots Windows Domain Controllers

The Hacker News · Analysis · Remediation

Severe Security Flaws Patched in Microsoft Dynamics 365 and Power Apps Web API

The Hacker News · Analysis · Remediation

Models

MBTSAD: Mitigating Backdoors in Language Models Based on Token Splitting and Attention Distillation

arXiv · Research · Models

Persistence of Backdoor-based Watermarks for Neural Networks: A Comprehensive Evaluation

arXiv · Research · Models

Backdoor Token Unlearning: Exposing and Defending Backdoors in Pretrained Language Models

arXiv · Research · Models

Mingling with the Good to Backdoor Federated Learning

arXiv · Research · Models

AI Security

AI Domination: Remote Controlling ChatGPT ZombAI Instances

Embrace The Red · Analysis · AI Security

Advisories in this window

  • 13 Packagist advisories · 5 high · 6 moderate · 2 low
  • 6 Go advisories · 1 critical · 4 high · 1 moderate
  • 4 npm advisories · 3 critical · 1 moderate
  • 1 crates.io advisory · 1 high
  • 1 PyPI advisory · 1 high
  • 1 Maven 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.