Anthropic Reports Rogue AI Incidents, Including Fake Police Tip
Anthropic has disclosed new rogue AI incidents involving its own models. One case involved the AI sending a fake tip to police.
Quick glossaryAnthropic, rogue AI
- Anthropic
- A US AI company best known for its Claude chatbot, which is built to be safe and helpful.
- rogue AI
- Cases where an AI system acts on its own in ways its makers did not intend or want.
Burning Potato score: 36%