The US Department of Commerce has rebranded its AI Safety Institute as the Center for AI Standards and Innovation (CAISI). This transition marks a shift in focus from general safety to addressing national security threats and reducing what officials describe as “burdensome and unnecessary regulation” in the international realm. Secretary of Commerce Howard Lutnick made the announcement on June 3rd, highlighting the agency’s efforts to “evaluate and enhance US innovation” and reinforce America’s leadership in global AI standards.
The AI Safety Institute was first introduced in 2023 during the Biden administration as part of a worldwide initiative aimed at developing best practices for government authorities to manage AI system risks. It established partnerships with prominent US AI firms, including OpenAI and Anthropic, to access new AI models and provide recommendations for improvements prior to their release. In the final days of Biden’s presidency in early 2025, the institute published draft guidelines addressing various AI risks, including the misuse of AI in generating biological weapons, as well as more prevalent issues such as child sexual abuse material (CSAM).
According to Lutnick’s comments, the new center will prioritize the assessment of “demonstrable risks” encompassing areas like cybersecurity, biosecurity, and chemical weapon threats. It will also explore the implications of “malign foreign influence” stemming from adversarial AI systems, which may encompass concerns such as DeepSeek, a Chinese large language model that significantly impacted the US AI sector earlier this year.
This initiative aligns with a broader effort by the Trump administration aimed at expediting the growth of American AI companies. Upon taking office, Trump revoked a Biden-era executive order mandating new safety measures for substantial AI systems and a report assessing the potential risks for US consumers and the job market. His executive orders have promoted the adoption of generative AI in sectors like education while advocating for coal as a power source for energy-intensive AI data centers. Furthermore, a current Republican budget proposal includes a 10-year freeze on state-level AI regulations, a measure that has faced opposition even from some within Trump’s own party.
