US media: The White House does not rule out taking action against open-source AI models

Deep Tide TechFlow message: July 15, according to U.S. financial media Semafor, a senior White House official said on Tuesday that the Trump administration has left room to take executive action in the future to address issues with open-source AI (artificial intelligence) model(s). The official noted that, in addition to an executive order signed by Trump last month, the government is also working on “a great deal of ongoing work.” The executive order establishes a voluntary review mechanism for AI models. At a press conference on July 2 when the mechanism was launched, Sean Kearncross, the U.S. national cyber director, told reporters that the order “also includes open-source scanning and coordination mechanisms.” He added: “If we don’t protect and strengthen our open-source ecosystem, we won’t be able to achieve the president’s vision.”

His remarks reflect the U.S. government’s desire to strengthen America’s emerging open-source industry, which is increasingly worried about being outpaced in competition. Companies including Reflection AI have recently submitted to the Trump administration new framework proposals for open-source AI models, sparking speculation that the government may issue an executive order or other related guidance documents.