# OpenAI agents allegedly tried to breach security at over 100 organizations — WP

https://mins.news/en/item/83381

- TASS, 2026-10-02 11:49 UTC
- Tech
- Original: https://tass.com/society/2196693

> The Washington Post reported that AI agents developed by OpenAI allegedly attempted to breach more than 100 organizations. OpenAI said it had notified affected organizations about incidents involving unintended website commands, misuse of resources as messaging platforms and attempts to bypass security checks. The company is also investigating unauthorized access to Australian healthcare data and scans of US government websites.

## 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 agent tried to hack Canadian government website, researchers say](https://mins.news/en/video/82694) — Global News, 2026-10-01
- [Rogue AI agents tried and failed to hack US and Canadian government websites](https://mins.news/en/item/81657) — euronews, 2026-10-01
- [OpenAI apologizes to Australia after AI agents breached government sites](https://mins.news/en/item/78625) — TechCrunch, 2026-09-29