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

Read on the web

Top developments

Google fixes new Chrome zero-day flaw exploited in attacks

BleepingComputer · Incident · Remediation

When The Impersonation Function Gets Used To Impersonate Users (Fortinet FortiWeb Auth. Bypass CVE-2025-64446)

watchTowr · Analysis · Remediation

CISA warns feds to fully patch actively exploited Cisco flaws

BleepingComputer · Advisory · Remediation

Fortinet’s delayed alert on actively exploited defect put defenders at a disadvantage

CyberScoop · Incident · Remediation

CISA warns of WatchGuard firewall flaw exploited in attacks

BleepingComputer · Incident · Remediation

SAST

SmartPoC: Generating Executable and Validated PoCs for Smart Contract Bug Reports

arXiv · Research · SAST

On the Security and Privacy of AI-based Mobile Health Chatbots

arXiv · Research · SAST

Multi-Agent Collaborative Fuzzing with Continuous Reflection for Smart Contracts Vulnerability Detection

arXiv · Research · DAST, SAST

Prompt Engineering vs. Fine-Tuning for LLM-Based Vulnerability Detection in Solana and Algorand Smart Contracts

arXiv · Research · SAST

SCA

Malicious NPM packages abuse Adspect redirects to evade security

BleepingComputer · Research · SCA

Over 67,000 Fake npm Packages Flood Registry in Worm-Like Spam Attack

The Hacker News · Incident · SCA

An LLM-based Quantitative Framework for Evaluating High-Stealthy Backdoor Risks in OSS Supply Chains

arXiv · Research · Models, SCA

Diffploit: Facilitating Cross-Version Exploit Migration for Open Source Library Vulnerabilities

arXiv · Research · SCA

DAST

SAINT: Service-level Integration Test Generation with Program Analysis and LLM-based Agents

arXiv · Research · DAST

Coverage-Guided Pre-Silicon Fuzzing of Open-Source Processors based on Leakage Contracts

arXiv · Research · DAST

Agentic SDLC

Agent READMEs: An Empirical Study of Context Files for Agentic Coding

arXiv · Research · Agentic SDLC

Vendor-Aware Industrial Agents: RAG-Enhanced LLMs for Secure On-Premise PLC Code Generation

arXiv · Research · Agentic SDLC

QLCoder: A Query Synthesizer For Static Analysis of Security Vulnerabilities

arXiv · Research · Agentic SDLC, SAST

Pre-Commit or Pay Later: The New Cost of AI-Era AppSec

Checkmarx · Analysis · Agentic SDLC

AI-BOM

AI Bill of Materials and Beyond: Systematizing Security Assurance through the AI Risk Scanning (AIRS) Framework

arXiv · Research · AI-BOM, SCA

Remediation

Microsoft Patch Tuesday addresses 63 defects, including one actively exploited zero-day

CyberScoop · Advisory · Remediation

Amazon pins Cisco, Citrix zero-day attacks to APT group

CyberScoop · Analysis · Remediation

When Attacks Come Faster Than Patches: Why 2026 Will be the Year of Machine-Speed Security

The Hacker News · Analysis · Remediation

Microsoft: Windows 10 KB5072653 OOB update fixes ESU install errors

BleepingComputer · Analysis · Remediation

Trust

Recap: Open Source Security Week in Belgium – Highlights from Ghent to Brussels

OpenSSF · Analysis · Trust

The Last Vote: A Multi-Stakeholder Framework for Language Model Governance

arXiv · Research · Trust

Models

SteganoBackdoor: Evading Data-Poisoning Defenses via Steganographic Backdoors

arXiv · Research · Models

Beyond Fixed and Dynamic Prompts: Embedded Jailbreak Templates for Advancing LLM Security

arXiv · Research · Models

Dynamic Black-box Backdoor Attacks on IoT Sensory Data

arXiv · Research · Models

ForgeDAN: An Evolutionary Framework for Jailbreaking Aligned Large Language Models

arXiv · Research · Models

AI Security

China’s ‘autonomous’ AI-powered hacking campaign still required a ton of human work

CyberScoop · Incident · AI Security

Scaling Patterns in Adversarial Alignment: Evidence from Multi-LLM Jailbreak Experiments

arXiv · Research · AI Security, Models

GRAPHTEXTACK: A Realistic Black-Box Node Injection Attack on LLM-Enhanced GNNs

arXiv · Research · AI Security, Models

Privacy-Preserving Prompt Injection Detection for LLMs Using Federated Learning and Embedding-Based NLP Classification

arXiv · Research · AI Security, Models

Advisories in this window

  • 68 npm advisories · 50 critical · 10 high · 7 moderate · 1 low
  • 20 Go advisories · 2 critical · 11 high · 4 moderate · 3 low
  • 10 PyPI advisories · 2 critical · 6 high · 1 moderate · 1 low
  • 6 Packagist advisories · 3 high · 2 moderate · 1 low
  • 6 Maven advisories · 2 high · 4 moderate
  • 3 crates.io advisories · 1 moderate · 2 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.