en.Wedoany.com Reported - OpenAI Group PBC, in a latest proposal, has put forward a different approach to regulating advanced artificial intelligence compared to the Trump administration's executive order. The company's document, "Democratic Governance of Frontier AI: A Blueprint for a Federal Framework," advocates for civilian agencies to oversee the safety of frontier AI, while the White House has designated the National Security Agency to assess potential risks. OpenAI prefers to entrust this task to the Center for AI Standards and Innovation (CAISI), a civilian scientific regulatory body under the National Institute of Standards and Technology within the Department of Commerce.
OpenAI CEO Sam Altman expressed support for the executive order on social media, stating that the U.S. should lead AI development by continuously building the best models, ensuring their safety, and placing cyber tools in the hands of trusted defenders, believing the new executive order strikes the right balance. Altman visited the White House that day for meetings with White House officials and key Democratic and Republican lawmakers as part of this week's agenda on discussions about the future of AI. According to reports, Altman is set to meet with Senator Bernie Sanders, who has been focusing on AI and its impact on the workforce, as well as discuss regulatory issues with Speaker Mike Johnson (R-La.) and Senate Minority Leader Chuck Schumer (D-N.Y.).
OpenAI executive Chris Lehane, in an interview with Politico, said the company is satisfied with the executive order but will seek to push President Trump to hand over the evaluation process to CAISI. Lehane noted that CAISI possesses the "sophisticated testing" required for such processes, and OpenAI, along with its competitor Anthropic PBC, has already shared information and built close relationships with CAISI. According to Lehane, another concern for OpenAI is the White House's plan to establish a "benchmarking" process for frontier AI models, which could lead to unclear questions about the extent of National Security Agency oversight over the models.
The executive order describes the benchmarking process as evaluating "the advanced cyber capabilities of AI models and determining the threshold at which an AI model should be designated as a 'covered frontier model' for the purposes of this order, and sharing such assessments with AI developers and researchers as appropriate." Lehane raised the question in the interview of when the capability threshold is reached and whether standards can be established to define this. These issues will be part of the discussions in future meetings.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
