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 7 of 2025 · Retrospective · Covers 4 February 2025 12:00 to 11 February 2025 12:00 UTC

Read on the web

Top developments

Apple fixes zero-day exploited in 'extremely sophisticated' attacks

BleepingComputer · Incident · Remediation

CISA tags Microsoft .NET and Apache OFBiz bugs as exploited in attacks

BleepingComputer · Incident · Remediation

Android security update includes patch for actively exploited vulnerability

CyberScoop · Incident · Remediation

Go Module Mirror served backdoor to devs for 3+ years

Ars Technica: Security · Research · SCA

Hackers Exploiting SimpleHelp RMM Flaws for Persistent Access and Ransomware

The Hacker News · Incident · Remediation

SAST

LLMs in Software Security: A Survey of Vulnerability Detection Techniques and Insights

arXiv · Research · SAST

Benchmarking Prompt Engineering Techniques for Secure Code Generation with GPT Models

arXiv · Research · SAST

Automated Exploit and Security Test Generation: A Systematic Literature Review

arXiv · Research · SAST

A Match Made in Heaven? AI-driven Matching of Vulnerabilities and Security Unit Tests

arXiv · Research · DAST, SAST

SCA

Pinning Is Futile: You Need More Than Local Dependency Versioning to Defend against Supply Chain Attacks

arXiv · Research · SCA

Sentient: Detecting APTs Via Capturing Indirect Dependencies and Behavioral Logic

arXiv · Research · SCA

Projecting the next decade of software supply chain security

CyberScoop · Analysis · SCA

From finding to fixing: GitHub Advanced Security integrates Endor Labs SCA

GitHub · Analysis · SCA

DAST

Certifying Language Model Robustness with Fuzzed Randomized Smoothing: An Efficient Defense Against Backdoor Attacks

arXiv · Research · DAST, Models

Agentic SDLC

Microsoft raises rewards for Copilot AI bug bounty program

BleepingComputer · Analysis · Agentic SDLC

OverThink: Slowdown Attacks on Reasoning LLMs

arXiv · Research · Agentic SDLC

Remediation

CISA orders agencies to patch Linux kernel bug exploited in attacks

BleepingComputer · Advisory · Remediation

Zyxel won’t patch newly exploited flaws in end-of-life routers

BleepingComputer · Advisory · Remediation

Progress Software Patches High-Severity LoadMaster Flaws Affecting Multiple Versions

The Hacker News · Advisory · Remediation

Microsoft shares workaround for Windows security update issues

BleepingComputer · Analysis · Remediation

Trust

TOCTOU Resilient Attestation for IoT Networks (Full Version)

arXiv · Research · Trust

Towards LLM Unlearning Resilient to Relearning Attacks: A Sharpness-Aware Minimization Perspective and Beyond

arXiv · Research · Models, Trust

AI Safety is Stuck in Technical Terms -- A System Safety Response to the International AI Safety Report

arXiv · Research · Trust

Enabling External Scrutiny of AI Systems with Privacy-Enhancing Technologies

arXiv · Research · Trust

Models

Revisiting the Auxiliary Data in Backdoor Purification

arXiv · Research · Models

DROP: Poison Dilution via Knowledge Distillation for Federated Learning

arXiv · Research · Models

Krum Federated Chain (KFC): Using blockchain to defend against adversarial attacks in Federated Learning

arXiv · Research · Models

Detection of Physiological Data Tampering Attacks with Quantum Machine Learning

arXiv · Research · Models

AI Security

Hacking Gemini's Memory with Prompt Injection and Delayed Tool Invocation

Embrace The Red · Analysis · AI Security

MELON: Provable Defense Against Indirect Prompt Injection Attacks in AI Agents

arXiv · Research · AI Security

KDA: A Knowledge-Distilled Attacker for Generating Diverse Prompts to Jailbreak LLMs

arXiv · Research · AI Security, Models

A Contemporary Survey of Large Language Model Assisted Program Analysis

arXiv · Research · AI Security

Advisories in this window

  • 91 npm advisories · 75 critical · 11 high · 5 moderate
  • 12 Maven advisories · 1 critical · 6 high · 5 moderate
  • 8 Packagist advisories · 1 critical · 3 high · 4 moderate
  • 7 Go advisories · 1 critical · 4 high · 2 moderate
  • 6 PyPI advisories · 5 high · 1 moderate
  • 4 crates.io advisories · 4 moderate
  • 1 Hex advisory · 1 low
  • 1 RubyGems advisory · 1 moderate
  • 1 Swift advisory · 1 high
  • 1 Multi-ecosystem 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.