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 21 of 2025 · Retrospective · Covers 13 May 2025 11:00 to 20 May 2025 11:00 UTC

Read on the web

Top developments

Fortinet fixes critical zero-day exploited in FortiVoice attacks

BleepingComputer · Incident · Remediation

Microsoft May 2025 Patch Tuesday fixes 5 exploited zero-days, 72 flaws

BleepingComputer · Advisory · Remediation

Ivanti fixes EPMM zero-days chained in code execution attacks

BleepingComputer · Analysis · Remediation

CISA tags recently patched Chrome bug as actively exploited

BleepingComputer · Advisory · Remediation

CVE-2025-42999: SAP NetWeaver Deserialization Vulnerability added to CISA's Known Exploited Vulnerabilities catalog

CISA · Advisory · SCA

SAST

GoLeash: Mitigating Golang Software Supply Chain Attacks with Runtime Policy Enforcement

arXiv · Research · SAST, SCA

Let the Trial Begin: A Mock-Court Approach to Vulnerability Detection using LLM-Based Agents

arXiv · Research · SAST

SCA

Malicious NPM package uses Unicode steganography to evade detection

BleepingComputer · Analysis · SCA

Malicious PyPI Package Posing as Solana Tool Stole Source Code in 761 Downloads

The Hacker News · Incident · SCA

Malicious PyPI Packages Exploit Instagram and TikTok APIs to Validate User Accounts

The Hacker News · Research · SCA

QUT-DV25: A Dataset for Dynamic Analysis of Next-Gen Software Supply Chain Attacks

arXiv · Research · SCA

DAST

Increase Red Team Operations 10X with Adversarial Exposure Validation

BleepingComputer · Research · DAST

AutoPentest: Enhancing Vulnerability Management With Autonomous LLM Agents

arXiv · Research · DAST, Remediation

Pen Testing for Compliance Only? It's Time to Change Your Approach

The Hacker News · Analysis · DAST

Agentic SDLC

Security Degradation in Iterative AI Code Generation -- A Systematic Analysis of the Paradox

arXiv · Research · Agentic SDLC

EVALOOOP: A Self-Consistency-Centered Framework for Assessing Large Language Model Robustness in Programming

arXiv · Research · Agentic SDLC, Models

LARGO: Latent Adversarial Reflection through Gradient Optimization for Jailbreaking LLMs

arXiv · Research · Agentic SDLC, Models

Remediation

Mozilla fixes Firefox zero-days exploited at hacking contest

BleepingComputer · Incident · Remediation

Wide-ranging Apple security update addresses over 30 vulnerabilities

CyberScoop · Analysis · Remediation

Samsung Patches CVE-2025-4632 Used to Deploy Mirai Botnet via MagicINFO 9 Exploit

The Hacker News · Analysis · Remediation

O2 UK patches bug leaking mobile user location from call metadata

BleepingComputer · Advisory · Remediation

Trust

Comparing Apples to Oranges: A Taxonomy for Navigating the Global Landscape of AI Regulation

arXiv · Research · Trust

Lara: Lightweight Anonymous Authentication with Asynchronous Revocation Auditability

arXiv · Research · Trust

Let's have a chat with the EU AI Act

arXiv · Research · Trust

LibVulnWatch: A Deep Assessment Agent System and Leaderboard for Uncovering Hidden Vulnerabilities in Open-Source AI Libraries

arXiv · Research · Trust

Models

AudioJailbreak: Jailbreak Attacks against End-to-End Large Audio-Language Models

arXiv · Research · Models

Causes and Consequences of Representational Similarity in Machine Learning Models

arXiv · Research · Models

PandaGuard: Systematic Evaluation of LLM Safety against Jailbreaking Attacks

arXiv · Research · Models

Does Low Rank Adaptation Lead to Lower Robustness against Training-Time Attacks?

arXiv · Research · Models

AI Security

From Assistants to Adversaries: Exploring the Security Risks of Mobile LLM Agents

arXiv · Research · AI Security

Improving Google A2A Protocol: Protecting Sensitive Data and Mitigating Unintended Harms in Multi-Agent Systems

arXiv · Research · AI Security, Models

Improving LLM Outputs Against Jailbreak Attacks with Expert Model Integration

arXiv · Research · AI Security, Models

CAPTURE: Context-Aware Prompt Injection Testing and Robustness Enhancement

arXiv · Research · AI Security

Advisories in this window

  • 290 npm advisories · 281 critical · 3 high · 5 moderate · 1 low
  • 17 Go advisories · 5 critical · 5 high · 6 moderate · 1 low
  • 12 Packagist advisories · 4 critical · 1 high · 5 moderate · 2 low
  • 9 Maven advisories · 2 critical · 2 high · 4 moderate · 1 low
  • 8 PyPI advisories · 5 high · 1 moderate · 2 low
  • 3 crates.io advisories · 1 high · 2 low
  • 2 Multi-ecosystem advisories · 1 critical · 1 moderate
  • 2 NuGet advisories · 1 high · 1 low
  • 1 GitHub Actions 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.