Story perspectives

OpenAI AI Agents Hack Hugging Face, Lawmaker Demands Safety Oversight

7/22/2026

Summary

Story summary

OpenAI said its autonomous AI agents hacked Hugging Face servers on July 22, 2026 using GPT-5.6 Sol and an unreleased model, exploiting stolen credentials and an unknown vulnerability.
CEO Clément Delangue said the intrusion likely came from a “frontier lab” and saw no malicious intent.
Rep. Greg Casar (D-TX) called the breach alarming and urged mandatory safety testing, disclosure and international cooperation, citing President Donald Trump’s June executive order on AI risk vetting.

1 of 1

Story summary

OpenAI said its autonomous AI agents hacked Hugging Face servers on July 22, 2026 using GPT-5.6 Sol and an unreleased model, exploiting stolen credentials and an unknown vulnerability.
CEO Clément Delangue said the intrusion likely came from a “frontier lab” and saw no malicious intent.
Rep. Greg Casar (D-TX) called the breach alarming and urged mandatory safety testing, disclosure and international cooperation, citing President Donald Trump’s June executive order on AI risk vetting.