# OpenAI reports more incidents of models acting deceptively

https://mins.news/en/item/62148

- Al Jazeera, 2026-09-17 06:15 UTC
- Tech
- Original: https://www.aljazeera.com/news/2026/9/17/openai-reports-more-incidents-of-models-acting-deceptively?traffic_source=rss

> OpenAI says internal training and testing uncovered additional cases in which AI models allegedly concealed mistakes or took unsanctioned actions. The ChatGPT maker will introduce a public framework for ongoing disclosures of unexpected or misaligned behaviour, while acknowledging that the industry has not yet adequately solved AI alignment and monitoring challenges.

## 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

- [OpenAI flags 6 new examples of 'concerning' AI behaviour](https://mins.news/en/item/62710) — CBC News, 2026-09-17
- [OpenAI flags concerning new AI behavior and vows closer monitoring](https://mins.news/en/video/62413) — The Associated Press, 2026-09-17
- [OpenAI finds more AI models acting deceptively during training](https://mins.news/en/item/62047) — CNN, 2026-09-17
- [OpenAI Reports New AI Safety Incidents and Sets Disclosure Process](https://mins.news/en/item/62025) — Bloomberg News, 2026-09-17