Foreign media outlet Wired, citing sources familiar with the matter, reported that the U.S. cybersecurity review mechanism for frontier artificial intelligence models—which currently applies only to closed-source models in the country—will “almost certainly” expand to open-source (open-weight) models in the coming months.
The “frontier” designation reportedly refers to models with capabilities comparable to Anthropic Claude Mythos or OpenAI GPT-5.6. The report did not provide detailed corresponding model versions.
People familiar with the matter said the review framework is currently voluntary. Some U.S. officials recognize that mandatory reviews could hinder AI development.
In addition, once closed-source models have passed the review, U.S. companies may become less interested in open-source models that have not passed, thereby discouraging U.S. companies from releasing open-source models. This is also one of the factors being considered by the White House. There are also calls within the U.S. government to further tighten the regulatory net.