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 39 of 2025 · Retrospective · Covers 16 September 2025 11:00 to 23 September 2025 11:00 UTC

Read on the web

Top developments

Google patches sixth Chrome zero-day exploited in attacks this year

BleepingComputer · Advisory · Remediation

Samsung patches zero-day security flaw used to hack into its customers’ phones

TechCrunch: Security · Advisory · Remediation

New attack on ChatGPT research agent pilfers secrets from Gmail inboxes

Ars Technica: Security · Incident · AI Security

PyPI invalidates tokens stolen in GhostAction supply chain attack

BleepingComputer · Incident · SCA

Microsoft Patches Critical Entra ID Flaw Enabling Global Admin Impersonation Across Tenants

The Hacker News · Advisory · Remediation

SAST

Security Evaluation of Android apps in budget African Mobile Devices

arXiv · Research · SAST

Detection of security smells in IaC scripts through semantics-aware code and language processing

arXiv · Research · SAST

STAFF: Stateful Taint-Assisted Full-system Firmware Fuzzing

arXiv · Research · DAST, SAST

AEAS: Actionable Exploit Assessment System

arXiv · Research · Remediation, SAST

SCA

Our plan for a more secure npm supply chain

GitHub · Analysis · SCA

NPM package caught using QR Code to fetch cookie-stealing malware

BleepingComputer · Analysis · SCA

Open Infrastructure is Not Free: A Joint Statement on Sustainable Stewardship

OpenSSF · Analysis · SCA

From Beginner to Builder: Your First Code Contribution

OpenSSF · Analysis · SCA

DAST

R1-Fuzz: Specializing Language Models for Textual Fuzzing via Reinforcement Learning

arXiv · Research · DAST

Revisiting Vulnerability Patch Localization: An Empirical Study and LLM-Based Solution

arXiv · Research · DAST, Remediation, SCA

Orion: Fuzzing Workflow Automation

arXiv · Research · DAST

Evaluating the Effectiveness of Coverage-Guided Fuzzing for Testing Deep Learning Library APIs

arXiv · Research · DAST

Agentic SDLC

New OpenSSF Guidance on AI Code Assistant Instructions

OpenSSF · Standard / policy · Agentic SDLC

httpjail

Simon Willison · Analysis · Agentic SDLC

SecureFixAgent: A Hybrid LLM Agent for Automated Python Static Vulnerability Repair

arXiv · Research · Agentic SDLC, Remediation, SAST

Gartner positions GitHub as a Leader in the 2025 Magic Quadrant for AI Code Assistants for the second year in a row

GitHub · Analysis · Agentic SDLC

Remediation

Fortra warns of max severity flaw in GoAnywhere MFT’s License Servlet

BleepingComputer · Analysis · Remediation

American Archive of Public Broadcasting fixes bug exposing restricted media

BleepingComputer · Analysis · Remediation

LLM-Driven SAST-Genius: A Hybrid Static Analysis Framework for Comprehensive and Actionable Security

arXiv · Research · Remediation, SAST

WatchGuard warns of critical vulnerability in Firebox firewalls

BleepingComputer · Analysis · Remediation

Trust

The AGNTCY Agent Directory Service: Architecture and Implementation

arXiv · Research · Trust

CoSAI Strategic Update September 2025

CoSAI · Analysis · Trust

Trustworthy and Confidential SBOM Exchange

arXiv · Research · SCA, Trust

Microsoft Purview innovations for your Fabric data: Unify data security and governance for the AI era

Microsoft · Analysis · Trust

Models

MER-Inspector: Assessing model extraction risks from an attack-agnostic perspective

arXiv · Research · Models

LLMZ+: Contextual Prompt Whitelist Principles for Agentic LLMs

arXiv · Research · AI Security, Models

Strategic Dishonesty Can Undermine AI Safety Evaluations of Frontier LLMs

arXiv · Research · Models

Modeling the Attack: Detecting AI-Generated Text by Quantifying Adversarial Perturbations

arXiv · Research · Models

AI Security

Why AI systems might never be secure

Simon Willison · Analysis · AI Security

Context Lineage Assurance for Non-Human Identities in Critical Multi-Agent Systems

arXiv · Research · AI Security

Design and Implementation of a Secure RAG-Enhanced AI Chatbot for Smart Tourism Customer Service: Defending Against Prompt Injection Attacks -- A Case Study of Hsinchu, Taiwan

arXiv · Research · AI Security

Privacy in Action: Towards Realistic Privacy Mitigation and Evaluation for LLM-Powered Agents

arXiv · Research · AI Security

Advisories in this window

  • 161 npm advisories · 149 critical · 4 high · 7 moderate · 1 low
  • 20 Maven advisories · 2 critical · 3 high · 15 moderate
  • 17 Go advisories · 6 high · 9 moderate · 2 low
  • 5 PyPI advisories · 1 critical · 3 high · 1 moderate
  • 3 NuGet advisories · 3 moderate
  • 3 Packagist advisories · 2 moderate · 1 low
  • 2 crates.io advisories · 1 high · 1 low
  • 1 Multi-ecosystem advisory · 1 critical
  • 1 RubyGems advisory · 1 low

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.