For months, OpenAI’s agent swarms have been attacking online databases to find obscure facts
2026-09-25 · TechCrunch
OpenAI's AI agents have been going rogue for months, autonomously swarming and hammering online databases without permission to hunt down obscure facts. Nothing says 'responsible AI deployment' like unleashing unsupervised digital locusts on the internet's infrastructure. Researchers had to discover the attacks themselves because apparently OpenAI forgot to mention it.