Weekly · Automated
AI Security — week of 2026-06-15
Automated digest. A language model running on my own machine selects and summarises these stories from public feeds each week. I do not review them item by item, so a headline or summary can be wrong. The labels, such as "corroborated", are generated too. Check the linked source before relying on anything here.
This week's stories
US restricts AI model access
The Hacker News · 2026-06-13 · corroborated · reporting
The US government has taken action to restrict access to AI models due to national security concerns.
Also: Simon Willison · BleepingComputer
AI-powered phishing
The Hacker News · 2026-06-12 · corroborated · reporting
AI-powered phishing attacks are becoming increasingly sophisticated and evasive.
Also: BleepingComputer · The Register Sec
LangGraph flaw exposes AI agents to RCE
The Hacker News · 2026-06-12 · single-source · reporting
A vulnerability in LangGraph has been discovered that exposes AI agents to remote code execution.
Also: The Hacker News · The Hacker News
AI agents vulnerable to phishing and RCE
The Hacker News · 2026-06-11 · corroborated · reporting
AI agents have been found to be vulnerable to phishing and remote code execution attacks.
Also: The Hacker News · The Hacker News
Anthropic revises Claude policy to add transparency
Simon Willison · 2026-06-11 · single-source · research
Anthropic has revised its Claude policy to add transparency and safeguards to its LLM development.
Also: Simon Willison · Simon Willison · The Hacker News
AI worm replicates autonomously
The Hacker News · 2026-06-09 · single-source · reporting
An autonomous AI worm has been discovered that can replicate itself without human intervention.
Hades PyPI attack poses supply chain risk
The Hacker News · 2026-06-09 · single-source · reporting
The Hades PyPI attack has highlighted the risk of supply chain attacks via poisoned packages.
Also: The Hacker News
CISA adds AI vulns to advisory list
CISA Advisories · 2026-06-08 · single-source · advisory
CISA has added AI vulnerabilities to its advisory list, urging remediation to avoid attacks.
China rebuilds botnets with AI datacenters at risk
The Register Sec · 2026-06-11 · single-source · reporting
China has been rebuilding its botnets, putting AI datacenters at risk of attack.
AI disrupts vuln management with speedup exploit development
The Hacker News · 2026-06-11 · single-source · reporting
AI has been found to disrupt vulnerability management by speeding up exploit development.
Also: The Hacker News
Claude Fable
Simon Willison · 2026-06-11 · single-source · research
Proactive AI debugs code
AI phishing
The Hacker News · 2026-06-08 · single-source · reporting
Overwhelming SOCs with high-volume alerts