The United States government will now be testing new artificial intelligence (AI) tools before they are released to the public.<\/p>\n
Tech firms Google, Microsoft and xAI have agreed to have the US Department of Commerce examine their models through the Center for AI Standards and Innovation (CAISI).<\/p>\n
The evaluations will cover \u201ctesting, collaborative research and best practice development related to commercial AI systems,\u201d the statement reads.<\/p>\n
CAISI will also be evaluating<\/a> \u201cdemonstrable risks\u201d associated with AI systems, such as cybersecurity, biosecurity and chemical weapons risks, its website reads.<\/p>\n \u201cIndependent, rigorous measurement science is essential to understanding frontier AI and its national security implications,\u201d said CAISI Director Chris Fall in a statement<\/a>. \u201cThese expanded industry collaborations help us scale our work in the public interest at a critical moment.<\/p>\n Microsoft said in a public statement<\/a> that the CAISI evaluations will help them stay ahead of risks, such as AI cyber attacks, for their AI model, Copilot.<\/p>\n Related<\/p>\n Shift in emphasis for Trump<\/p>\n The announcement that AI companies will be evaluated is a departure for President Trump, who has long argued that excessive regulation of AI systems could hurt the country\u2019s innovation and could allow China to gain an advantage.<\/p>\n In March, Trump released his AI National Policy Framework, which says that the United States will \u201cremove barriers to innovation\u201d and \u201caccelerate\u201d the deployment of AI across various sectors.<\/p>\n It also said that Congress will not create \u201cany few federal rulemaking bodies to regulate AI,\u201d but instead get existing regulatory bodies and experts in specific domains to examine the models.<\/p>\n CAISI has already conducted 40 evaluations of other models, including on some \u201cstate-of-the-art models that remain unreleased,\u201d but did not specify which models it is referring to.<\/p>\n “Renegotiation” of existing agreements<\/p>\n OpenAI and Anthropic signed agreements for these evaluations<\/a> in 2024 under former President Joe Biden. CAISI said existing agreements had been \u201crenegotiated,\u201d but did not elaborate on what had changed.<\/p>\n