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 12 of 2025 · Retrospective · Covers 11 March 2025 11:00 to 18 March 2025 11:00 UTC

Read on the web

Top developments

Apple fixes WebKit zero-day exploited in ‘extremely sophisticated’ attacks

BleepingComputer · Incident · Remediation

Microsoft patches 57 vulnerabilities, including 6 zero-days

CyberScoop · Advisory · Remediation

GitHub Action tj-actions/changed-files supply chain attack: everything you need to know

Wiz · Incident · SCA

CVE-2025-24201: Apple Multiple Products WebKit Out-of-Bounds Write Vulnerability added to CISA's Known Exploited Vulnerabilities catalog

CISA · Advisory · Agentic SDLC

Microsoft patches Windows Kernel zero-day exploited since 2023

BleepingComputer · Advisory · Remediation

SAST

CASTLE: Benchmarking Dataset for Static Code Analyzers and LLMs towards CWE Detection

arXiv · Research · SAST

SCA

GHSA-mrrh-fwg8-r2c3 / CVE-2025-30066: tj-actions changed-files through 45.0.7 allows remote attackers to discover secrets by reading actions logs.

GitHub Advisory Database · Advisory · SCA

Large enterprises scramble after supply-chain attack spills their secrets

Ars Technica: Security · Incident · SCA

North Korean Lazarus hackers infect hundreds via npm packages

BleepingComputer · Analysis · SCA

Malicious PyPI Packages Stole Cloud Tokens—Over 14,100 Downloads Before Removal

The Hacker News · Incident · SCA

DAST

Pentesters: Is AI Coming for Your Role?

The Hacker News · Analysis · DAST

Agentic SDLC

Auspex: Building Threat Modeling Tradecraft into an Artificial Intelligence-based Copilot

arXiv · Research · Agentic SDLC

GitHub for Beginners: Essential features of GitHub Copilot

GitHub · Analysis · Agentic SDLC

Remediation

Microsoft March 2025 Patch Tuesday fixes 7 zero-days, 57 flaws

BleepingComputer · Advisory · Remediation

Cisco IOS XR vulnerability lets attackers crash BGP on routers

BleepingComputer · Analysis · Remediation

MIP against Agent: Malicious Image Patches Hijacking Multimodal OS Agents

arXiv · Research · Remediation

GitLab patches critical authentication bypass vulnerabilities

BleepingComputer · Analysis · Remediation

Trust

Augmenting Software Bills of Materials with Software Vulnerability Description: A Preliminary Study on GitHub

arXiv · Research · SCA, Trust

The Role of Legal Frameworks in Shaping Ethical Artificial Intelligence Use in Corporate Governance

arXiv · Research · Trust

Ethical Implications of AI in Data Collection: Balancing Innovation with Privacy

arXiv · Research · Trust

Closing the Chain: How to reduce your risk of being SolarWinds, Log4j, or XZ Utils

arXiv · Research · SCA, Trust

Models

Optimizing ML Training with Metagradient Descent

arXiv · Research · Models

ProDiF: Protecting Domain-Invariant Features to Secure Pre-Trained Models Against Extraction

arXiv · Research · Models

MirrorShield: Towards Universal Defense Against Jailbreaks via Entropy-Guided Mirror Crafting

arXiv · Research · Models

D-ADD: An Effective Plug-In for Defending Against Model Stealing

arXiv · Research · Models, SCA

AI Security

Prompt Injection Detection and Mitigation via AI Multi-Agent NLP Frameworks

arXiv · Research · AI Security

Sneaky Bits: Advanced Data Smuggling Techniques (ASCII Smuggler Updates)

Embrace The Red · Analysis · AI Security

CyberLLMInstruct: A Pseudo-malicious Dataset Revealing Safety-performance Trade-offs in Cyber Security LLM Fine-tuning

arXiv · Research · AI Security

Probing Latent Subspaces in LLM for AI Security: Identifying and Manipulating Adversarial States

arXiv · Research · AI Security, Models

Advisories in this window

  • 349 npm advisories · 342 critical · 3 high · 3 moderate · 1 low
  • 13 Go advisories · 3 critical · 4 high · 6 moderate
  • 9 Packagist advisories · 1 high · 6 moderate · 2 low
  • 9 PyPI advisories · 1 critical · 3 high · 5 moderate
  • 7 Maven advisories · 1 high · 5 moderate · 1 low
  • 7 RubyGems advisories · 5 critical · 2 high
  • 4 crates.io advisories · 2 high · 1 moderate · 1 low
  • 3 NuGet advisories · 1 high · 2 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.