OpenAI Delays Release of Open Model for Further Safety Testing
OpenAI CEO Sam Altman announced on Friday that the company is postponing the release of its highly anticipated open model. Originally scheduled for release next week, the open model has been pushed back indefinitely to allow for additional safety testing.
Altman explained the decision in a post on X, stating, “We need time to run additional safety tests and review high-risk areas. We are not yet sure how long it will take us. While we trust the community will build great things with this model, once weights are out, they can’t be pulled back. This is new for us and we want to get it right.”
OpenAI’s open model release has been eagerly anticipated by the AI community, alongside the expected release of GPT-5. Unlike GPT-5, the open model will be freely available for developers to download and run locally. This launch is crucial for OpenAI to demonstrate its position as a leading AI lab in Silicon Valley, amidst increasing competition from other tech giants.
The delay means developers will have to wait a little longer to try the first open model released by OpenAI in years. JS previously reported that the open model is expected to have similar reasoning capabilities to the company’s o-series of models and aims to be best-in-class compared to other open models.
Earlier this week, Chinese AI startup Moonshot AI launched Kimi K2, a one-trillion-parameter open AI model that outperforms OpenAI’s GPT-4.1 on several benchmarks, adding to the competitive landscape of open AI models.
In June, Altman hinted at the remarkable capabilities of the open model but emphasized the need for more time to ensure a high-quality release. Aidan Clark, OpenAI’s VP of research leading the open model team, echoed this sentiment, stating, “Capability wise, we think the model is phenomenal — but our bar for an open source model is high and we think we need some more time to make sure we’re releasing a model we’re proud of along every axis.”
There have been discussions within OpenAI about enabling the open model to connect to the company’s cloud-hosted AI models for complex queries, although it remains unclear if these features will be included in the final release.
