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 25 of 2025 · Retrospective · Covers 10 June 2025 11:00 to 17 June 2025 11:00 UTC

Read on the web

Top developments

Microsoft June 2025 Patch Tuesday fixes exploited zero-day, 66 flaws

BleepingComputer · Advisory · Remediation

Zero-click AI data leak flaw uncovered in Microsoft 365 Copilot

BleepingComputer · Incident · Agentic SDLC, AI Security, Remediation

Fog ransomware attack uses unusual mix of legitimate and open-source tools

BleepingComputer · Incident · DAST

Password-spraying attacks target 80,000 Microsoft Entra ID accounts

BleepingComputer · Analysis · DAST

An Introduction to Google’s Approach to AI Agent Security

Simon Willison · Research · AI Security

SAST

Identifying Helpful Context for LLM-based Vulnerability Repair: A Preliminary Study

arXiv · Research · Remediation, SAST

AI-Based Software Vulnerability Detection: A Systematic Literature Review

arXiv · Research · SAST

Expert-in-the-Loop Systems with Cross-Domain and In-Domain Few-Shot Learning for Software Vulnerability Detection

arXiv · Research · SAST

ZTaint-Havoc: From Havoc Mode to Zero-Execution Fuzzing-Driven Taint Inference

arXiv · Research · DAST, SAST

SCA

PyPI, npm, and AI Tools Exploited in Malware Surge Targeting DevOps and Cloud Environments

The Hacker News · Incident · SCA

Information-theoretic Estimation of the Risk of Privacy Leaks

arXiv · Research · SCA

Scalable Software Testing in Fast Virtual Platforms: Leveraging SystemC, QEMU and Containerization

arXiv · Research · SCA

GUAC 1.0 is Now Available

OpenSSF · Release · SCA

DAST

Recognition Without Mitigation: Ethical Frameworks in Autonomous Offensive-LLM Agent Research

arXiv · Research · DAST

Agentic SDLC

Reasoning as a Resource: Optimizing Fast and Slow Thinking in Code Generation Models

arXiv · Research · Agentic SDLC

Risks & Benefits of LLMs & GenAI for Platform Integrity, Healthcare Diagnostics, Financial Trust and Compliance, Cybersecurity, Privacy & AI Safety: A Comprehensive Survey, Roadmap & Implementation Blueprint

arXiv · Research · Agentic SDLC

How the GitHub billing team uses the coding agent in GitHub Copilot to continuously burn down technical debt

GitHub · Analysis · Agentic SDLC

Remediation

Apple Zero-Click Flaw in Messages Exploited to Spy on Journalists Using Paragon Spyware

The Hacker News · Incident · Remediation

Dems want watchdog study of two troubled federally-funded vulnerability tracking initiatives

CyberScoop · Research · Remediation

Found in the wild: 2 Secure Boot exploits. Microsoft is patching only 1 of them.

Ars Technica: Security · Research · Remediation

Microsoft shares temp fix for Outlook crashes when opening emails

BleepingComputer · Analysis · Remediation

Trust

Tech Talk Recap | CRA-Ready: How Open Source Projects Can Prepare for the EU Cyber Resilience Act

OpenSSF · Standard / policy · Trust

Step-by-Step Reasoning Attack: Revealing 'Erased' Knowledge in Large Language Models

arXiv · Research · Models, Trust

Data-Centric Safety and Ethical Measures for Data and AI Governance

arXiv · Research · Trust

Models

Theoretically Unmasking Inference Attacks Against LDP-Protected Clients in Federated Vision Models

arXiv · Research · Models

Unlearning-Enhanced Website Fingerprinting Attack: Against Backdoor Poisoning in Anonymous Networks

arXiv · Research · Models

Thought Crime: Backdoors and Emergent Misalignment in Reasoning Models

arXiv · Research · Models

Jailbreak Transferability Emerges from Shared Representations

arXiv · Research · Models

AI Security

100% effective

Simon Willison · Analysis · AI Security

We Should Identify and Mitigate Third-Party Safety Risks in MCP-Powered Agent Systems

arXiv · Research · AI Security

The lethal trifecta for AI agents: private data, untrusted content, and external communication

Simon Willison · Analysis · AI Security

Shadow AI: Examples, Risks, and 8 Ways to Mitigate Them

Mend.io · Research · AI Security

Advisories in this window

  • 61 npm advisories · 56 critical · 3 high · 2 moderate
  • 28 Maven advisories · 2 critical · 17 high · 9 moderate
  • 20 PyPI advisories · 1 critical · 3 high · 13 moderate · 3 low
  • 20 Packagist advisories · 1 critical · 4 high · 14 moderate · 1 low
  • 9 Go advisories · 1 critical · 2 high · 4 moderate · 2 low
  • 2 RubyGems advisories · 1 critical · 1 high
  • 2 crates.io advisories · 2 moderate
  • 1 NuGet 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.