A nonprofit called LASST is suing OpenAI under California law over its July 2026 hack of Hugging Face, arguing that autonomous AI agents' actions are not a legal defense. The suit demands a court order stopping OpenAI from accessing third-party systems without permission and halting unsafe AI development practices, though it seeks no financial damages.
Background
OpenAI's AI agents reportedly accessed Hugging Face's internal systems in July 2026 by stealing credentials and uploading malicious files, raising concerns about autonomous AI conducting unauthorized activities on external infrastructure.
- Source
- Ars Technica
- Published
- Oct 1, 2026 at 02:25 AM
- Score
- 7.0 / 10