WSJ reports OpenAI scrapped GPT-6.1 Astra over safety concerns

The Wall Street Journal reported that OpenAI scrapped the release of GPT-6.1 Astra over safety concerns raised by researchers during internal testing, with the model due to debut in ChatGPT and Codex in October, according to a September 28th, 2026 X post by Leo (@synthwavedd) that shared the Journal headline. OpenAI's public materials document GPT-6 Astra, introduced September 3rd, but no GPT-6.1 release; a September 1st safety update said Astra met OpenAI's "Critical" threshold for cybersecurity capability, prompting the company to delay parts of its development and release before launching it two days later with limits on its most advanced cybersecurity capabilities. The reported October debut would have put GPT-6.1 Astra in ChatGPT and Codex, though public materials do not specify what would have changed in its capabilities, API terms or access.

WSJ reports OpenAI scrapped GPT-6.1 Astra over safety concerns The model was due to debut in ChatGPT and Codex in October, after researchers raised concerns during internal testing, the Journal reports. By Ryan Merket https://runtimewire.com/author/ryan-merket · Published Primary source: X https://x.com/synthwavedd/status/2104694911631007801 Why it matters The post points to a safety question with a real precedent: OpenAI delayed parts of Astra's development before releasing it. But the public record confirms GPT-6 Astra, not an October GPT-6.1 plan or its cancellation. A Wall Street Journal report by Maxwell Zeff https://www.wsj.com/tech/ai/openai-chatgpt-model-release-cancel-safety-5a2f9f42 says OpenAI is scrapping the release of GPT-6.1 Astra over safety concerns raised by researchers during internal testing. The model was due to debut in ChatGPT and Codex in October, the Journal reports. That confirms the report behind an X post by Leo @synthwavedd https://x.com/synthwavedd/status/2104694911631007801 , which shared an image of the Journal headline. The post was published on September 28th, 2026. OpenAI's public product and API materials identify GPT-6 Astra https://runtimewire.com/models/native-openai/gpt-6-astra-576a599a64cb3dde , introduced on September 3rd, but do not document a GPT-6.1 release. The company described Astra as a model for complex reasoning, coding, computer use, research and professional work, with availability through ChatGPT plans, the API, Microsoft Azure and Amazon Bedrock. Its API catalog https://developers.openai.com/api/docs/models/gpt-6-astra lists a 1.05-million-token context window, a maximum output of 128,000 tokens and rates of $10 per million input tokens and $50 per million output tokens. OpenAI had already disclosed safety concerns around Astra. In a September 1st safety update https://openai.com/index/path-to-astra/ , the company said the model met its "Critical" threshold for cybersecurity capability, meaning it could identify previously unknown flaws and develop exploits in well-protected systems without a person guiding each step. OpenAI said it delayed parts of Astra's development and release while strengthening safeguards. It released Astra two days later, saying the protections were sufficient for deployment, while limiting access to the model's most advanced cybersecurity capabilities. The API changelog https://developers.openai.com/api/docs/changelog records GPT-6 Sol https://runtimewire.com/models/native-openai/gpt-6-sol-7388db16debdff68 and GPT-6 Luna https://runtimewire.com/models/native-openai/gpt-6-luna-9a7e60cd1c7bb9d5 releases on September 22nd, alongside Astra in the GPT-6 family. Those public records describe available models, while the Journal's report concerns a planned GPT-6.1 release and its cancellation. The reported October debut would have put the model in ChatGPT and Codex; the available public materials do not specify what would have changed in its capabilities, API terms or access.