US frontier labs face criticism for pushing ban on open-source AI

An article published by Crypto Briefing reports that US frontier AI laboratories are facing criticism for pushing a ban on open-source AI. The piece frames this as an ongoing debate over open-weight models, set against competition from Chinese models such as GLM-5.2. The excerpt provides no additional named labs or critics.

According to the report, the Trump administration has considered sanctions on specific Chinese AI systems, but has ruled out a comprehensive ban on open-weight models. In June, Anthropic received US export-control orders limiting international access to its Fable 5 and Mythos 5 models. The excerpt gives no further details on those orders, such as which countries or customers are affected, and does not state whether other US labs received similar orders. It also does not specify what the export-control orders cover beyond access to the two models.

The report frames these developments as part of a broader argument about open-weight models and competitiveness. It suggests that the criticism of US frontier labs stems from the combination of their position on open-source AI and the emergence of capable Chinese open models like GLM-5.2. The administration’s decision not to impose a comprehensive ban on open-weight models is presented as a notable limit on current US policy. The article highlights a contrast between two policy approaches: targeted sanctions and export controls, which are either considered or in effect, and a general prohibition on open-source AI, which the administration has ruled out. The excerpt does not state whether the export-control orders imposed on Anthropic were related to any ban the labs are said to be pushing.