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 48 of 2025 · Retrospective · Covers 18 November 2025 12:00 to 25 November 2025 12:00 UTC

Read on the web

Top developments

Shai-Hulud 2.0 Supply Chain Attack: 25K+ Repos Exposing Secrets

Wiz · Incident · SCA

CISA warns Oracle Identity Manager RCE flaw is being actively exploited

BleepingComputer · Incident · Remediation

CISA gives govt agencies 7 days to patch new Fortinet flaw

BleepingComputer · Advisory · Remediation

Coalition for Secure AI Releases Two Actionable Frameworks for AI Model Signing and Incident Response

CoSAI · Incident · SCA

‘PlushDaemon’ hackers hijack software updates in supply-chain attacks

BleepingComputer · Incident · SCA

SAST

BASICS: Binary Analysis and Stack Integrity Checker System for Buffer Overflow Mitigation

arXiv · Research · SAST

LLM-CSEC: Empirical Evaluation of Security in C/C++ Code Generated by Large Language Models

arXiv · Research · SAST

Towards Classifying Benign And Malicious Packages Using Machine Learning

arXiv · Research · SAST, SCA

SCA

The slow rise of SBOMs meets the rapid advance of AI

CyberScoop · Analysis · Agentic SDLC, SCA

KubeCon Keynote Recap: “Supply Chain Reaction” and Why the OSPS Baseline Matters More Than Ever

OpenSSF · Analysis · SCA

Tech Talk Recap: Simplifying DevSecOps in Air-Gapped Environments with Zarf

OpenSSF · Analysis · SCA

Checkmarx Zero Takes Down Malicious “Prettier” Alternative Found In VSCode Marketplace

Checkmarx · Analysis · SCA

DAST

MASTEST: A LLM-Based Multi-Agent System For Testing RESTful APIs

arXiv · Research · DAST

ThreadFuzzer: Fuzzing Framework for Thread Protocol

arXiv · Research · DAST

Hiding in the AI Traffic: Abusing MCP for LLM-Powered Agentic Red Teaming

arXiv · Research · DAST

The Cloudflare Outage May Be a Security Roadmap

Krebs on Security · Analysis · DAST

Agentic SDLC

DualGauge: Automated Joint Security-Functionality Benchmarking of Specification-Only Code Generation by LLMs and Coding Agents

arXiv · Research · Agentic SDLC

Multi-Agent Code Verification via Information Theory

arXiv · Research · Agentic SDLC, Remediation

Critics scoff after Microsoft warns AI feature can infect machines and pilfer data

Ars Technica: Security · Analysis · Agentic SDLC, AI Security

Agents built into your workflow: Get Security Copilot with Microsoft 365 E5

Microsoft · Analysis · Agentic SDLC

Remediation

ShadowPad Malware Actively Exploits WSUS Vulnerability for Full System Access

The Hacker News · Analysis · Remediation

Nvidia confirms October Windows updates cause gaming issues

BleepingComputer · Analysis · Remediation

Grafana Patches CVSS 10.0 SCIM Flaw Enabling Impersonation and Privilege Escalation

The Hacker News · Advisory · Remediation

New SonicWall SonicOS flaw allows hackers to crash firewalls

BleepingComputer · Analysis · Remediation

Trust

Normative active inference: A numerical proof of principle for a computational and economic legal analytic approach to AI governance

arXiv · Research · Trust

Constant-Size Cryptographic Evidence Structures for Regulated AI Workflows

arXiv · Research · Trust

MAIF: Enforcing AI Trust and Provenance with an Artifact-Centric Agentic Paradigm

arXiv · Research · Trust

Model Context Protocol SEP-1865: MCP Apps - Interactive User Interfaces for MCP

Model Context Protocol · Proposal · AI Security, Trust

Models

Natural emergent misalignment from reward hacking

Anthropic · Research · Models

Cross-LLM Generalization of Behavioral Backdoor Detection in AI Agent Supply Chains

arXiv · Research · AI Security, Models

Are Neuro-Inspired Multi-Modal Vision-Language Models Resilient to Membership Inference Privacy Leakage?

arXiv · Research · Models

FedPoisonTTP: A Threat Model and Poisoning Attack for Federated Test-Time Personalization

arXiv · Research · Models

AI Security

Prompt Fencing: A Cryptographic Approach to Establishing Security Boundaries in Large Language Model Prompts

arXiv · Research · AI Security

Mitigating prompt injections in browser use

Anthropic · Analysis · AI Security

Semantics as a Shield: Label Disguise Defense (LDD) against Prompt Injection in LLM Sentiment Classification

arXiv · Research · AI Security

TASO: Jailbreak LLMs via Alternative Template and Suffix Optimization

arXiv · Research · AI Security, Models

Advisories in this window

  • 621 npm advisories · 610 critical · 4 high · 6 moderate · 1 low
  • 19 Go advisories · 2 critical · 7 high · 7 moderate · 3 low
  • 19 Packagist advisories · 1 high · 12 moderate · 6 low
  • 9 PyPI advisories · 2 critical · 3 high · 3 moderate · 1 low
  • 6 Maven advisories · 2 critical · 1 high · 2 moderate · 1 low
  • 1 crates.io 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.