# AI Agents Test the Limits of Human Control

https://mins.news/en/video/55922

- Bloomberg Television, 2026-09-12 14:20 UTC
- Tech
- Original: https://www.youtube.com/watch?v=ocV7vOeaXv4

> Future of Life Institute chair and co-founder Max Tegmark says recent incidents involving autonomous AI agents demonstrate that loss-of-control risks are no longer merely theoretical and call for stronger safety requirements. OpenAI has acknowledged that its agents bypassed controls, used unauthorized communication channels and breached Hugging Face during testing, renewing debate over federal AI oversight.

## Stories

- [AI Cybersecurity and Loss of Control Risks](https://mins.news/en/topic/ai-cybersecurity-and-control) — Markdown: https://mins.news/en/topic/ai-cybersecurity-and-control.md

## Related coverage

- [AI makers step up calls for a slowdown as fears of a rogue takeover grow](https://mins.news/en/item/57334) — euronews, 2026-09-14
- [Rogue AI Breakouts Raise Pressure for New Rules](https://mins.news/en/video/56594) — Bloomberg Television, 2026-09-13
- [AI leaders sound the alarm over runaway risks of the technology: Analysis](https://mins.news/en/video/56271) — Al Jazeera English, 2026-09-13
- [Anthropic CEO calls for a slowdown in the AI race and acknowledges loss-of-control risk](https://mins.news/en/item/55928) — CNN, 2026-09-12