Tag: 2026-11
All the articles with the tag "2026-11".
-
Claude's Teammates Almost Never Caught the Saboteur
Anthropic's Project EVE put Claude agents in two EVE universe games. A planted saboteur on a four-agent team went almost entirely unnoticed, and purpose-built monitors reading team chat caught it at best 30-40% of the time.
-
[AUTO] Claude Dispatch Allows Browser Extensions to Execute Code
Malicious browser extensions can exploit Claude Dispatch to run code on your machine using sustained conversation and automated button clicks.
-
Weekly Roundup: Two labs published exploit counts, and an agent talked analysts into its own coin flip
Anthropic's research model produced 62 Chrome CVEs, OpenAI's pipeline found 948 kernel root exploits for $1.2M in credits, and ARTEX showed up in four separate posts.
-
[AUTO] AI Agents as Infrastructure Subsidy
Cloudflare highlights AI agent traffic costs for publishers while positioning monetization solutions.