Restrictions on advanced proprietary AI models are increasing enterprise adoption of open-source and open-weight alternatives that offer greater deployment flexibility and control.
Government restrictions in the US on access to AI technologies have resulted in more interest in open-source and open-weight AI models, particularly as businesses seek alternatives that are less vulnerable to government controls and platform restrictions.
This comes after restrictions were placed on Anthropic’s Mythos 5 and Fable 5 models, followed by OpenAI making GPT-5.6 available only through a government-approved preview programme. The reason why open-source and open-weight models are preferred is that they can be downloaded, modified, and deployed locally, which means more control over the AI system than proprietary models.
Last month, the Trump administration instructed Anthropic to block non-Americans from accessing its most advanced AI models. Instead of implementing a user verification process, Anthropic opted out of offering those models to everyone. OpenAI has decided to release GPT-5.6 via the government-approved preview programme for selected customers.
Oren Michels, co-founder and chief executive of Barndoor AI, said, “If everything you need to do has to be on a specific frontier model, that makes whatever you’re building a whole lot less reliable when access is suddenly withdrawn.”
Haitham Mengad, co-founder of AI music startup Stems Labs, said the withdrawal of Anthropic’s Fable model demonstrated how dependent some developers had become on closed AI platforms.
The increasing use of open models is also impacting the view of Chinese AI. According to Mengad, open Chinese models do not pose significant security risks because, once downloaded and run locally, developers retain control over their enterprise data and how the models are used.
