Tag: AI agents
Spain logs first breach blamed on an agent acting alone
An agent found a flaw, logged in and edited personal records without anyone directing it there.
OpenAI agents flooded RubyGems with thousands of fake packages
New research ties May's RubyGems junk-package flood to a swarm of autonomous OpenAI agents.
Anthropic finds a fourth case of Claude breaching real systems
A widened transcript scan turned up another incident of a model reaching a third party.
Dormant German wiki hosted 18,000 secret posts from OpenAI agents
Researchers say OpenAI agents left roughly 18,000 posts on a dormant German wiki while coordinating on timed tasks.
Poisoned Git configs make AI coding agents run attacker commands
Manifold finds eight flaws where poisoned Git settings make seven AI coding agents run attacker commands.
OpenAI details how rogue agents escaped and hit Hugging Face
OpenAI's report on the Hugging Face breach details how AI agents coordinated through internal systems to escape.
Agentic hackers breached Taiwan agencies with self-correcting AI tool
Israeli firm Dream documents what appears to be the first autonomous AI attack on a government.
Poisoned logs turn trusted platforms into agent hijackers
Tenet Security's Ghostjacking demo shows attackers planting instructions in logs that AI agents read and execute, targeting Cloudflare, Datadog, and…