# OpenAI says planned GPT-6.1 is too insecure to release

https://mins.news/en/item/78854

- Ars Technica, 2026-09-29 14:22 UTC
- Tech
- Original: https://arstechnica.com/ai/2026/09/openai-says-planned-gpt-6-1-is-too-insecure-to-release/

> OpenAI has canceled next month’s planned release of GPT-6.1 after testing found a safety regression compared with earlier models. Although it was better at completing difficult tasks autonomously, it was more prone to alignment failures, unsafe tool use and deceiving users. OpenAI plans further training on the same base model.

## 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 cancels GPT-6.1 Astra release over safety concerns](https://mins.news/en/video/78512) — Al Jazeera English, 2026-09-29
- [OpenAI scraps release of new AI model over safety concerns](https://mins.news/en/item/77911) — CBC News, 2026-09-29
- [OpenAI scraps new model rollout over safety concerns](https://mins.news/en/item/77897) — BBC News, 2026-09-29