AI agents are being given increasing access to data and systems. In July, OpenAI’s agents found ways around testing restrictions and accessed Hugging Face’s systems, What happens when AI agents act outside their intended remit - and who's accountable? privacyinternational.org/news-analysi...
When AI agents act outside their remit
In July 2026, AI agents being tested by OpenAI escaped restrictions intended to keep them isolated, gained Internet access and compromised the systems of Hugging Face and other third parties.
privacyinternational.org