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

Daily, Tuesday 1 September 2026 · Retrospective · Covers 31 August 2026 00:00 to 1 September 2026 00:00 UTC

Read on the web

Top developments

Aurora Ransomware Operators Use Cursor AI in Attacks Against 10 Targets

The Hacker News · Incident · Agentic SDLC

CVE-2026-82078: PaperCut NG/MF Unsafe Reflection Vulnerability added to CISA's Known Exploited Vulnerabilities catalog

CISA · Advisory

CVE-2026-81578: PaperCut NG/MF Missing Authentication for Critical Function Vulnerability added to CISA's Known Exploited Vulnerabilities catalog

CISA · Advisory

SCA

sbom-unifier: Integration Framework for Heterogeneous SBOMs

arXiv · Research · Remediation, SCA

Agentic SDLC

Beyond the Payload: How User Invocation Shapes Coding Agent Vulnerability to Repository Poisoning

arXiv · Research · Agentic SDLC

ECLIPSE: Self-Evolving Stealthy Prompt Injection Attack against Long-Horizon Agentic Systems

arXiv · Research · Agentic SDLC, AI Security

Memory as Infrastructure: Reliability Engineering for Persistent Agent Memory in Months-Long LLM-Assisted Development

arXiv · Research · Agentic SDLC

Trust

Delegation Without Trust: An Empirical Gap Analysis of Identity, Authorization, and Runtime Governance in Multi-Agent LLM Systems

arXiv · Research · AI Security, Trust

KORD: Breaking the Key-Generation Bottleneck in Dealerless FSS via Protocol--Hardware Co-Design

arXiv · Research · Trust

UN Global Dialogue Co-Leads Issue Call to Action on Safe, Secure, and Trustworthy AI

Partnership on AI · Analysis · Trust

Models

Does Reasoning Mitigate Backdoor Attacks? A Neuro-Symbolic Perspective

arXiv · Research · Models

Context Inference Attacks Without Jailbreaks

arXiv · Research · Models

Do VLMs Share Safety Neurons Across Modalities?

arXiv · Research · Models

AI Security

Attesting Outputs and Delegation Ancestry in Multi-Agent AI Systems

arXiv · Research · AI Security

Will the User Ever Know? Covert Indirect Prompt Injection Attacks on Tool-Using LLM Agents

arXiv · Research · AI Security

Can MCP Clients Decide What to Do After Failure? A Result-Only Actionability Audit

arXiv · Research · AI Security

Advisories in this window

  • 4385 npm advisories · 4381 critical · 1 high · 3 moderate
  • 4 Packagist advisories · 3 high · 1 moderate
  • 2 PyPI advisories · 2 critical

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.