Trump Calls Off AI Safety Executive Order Citing Innovation Concerns

President Donald Trump abruptly postponed a highly anticipated White House ceremony intended to formalize a sweeping executive order on artificial intelligence safety. The directive, which aimed to establish a federal vetting framework for advanced AI models, was withdrawn just hours before a scheduled press conference with leading technology executives. According to reports, President Trump expressed concerns that certain provisions within the draft order could act as a "blocker" to American innovation, potentially impeding the nation's technological lead. This decision underscores an ongoing internal debate within the administration, balancing national security considerations against the imperative to foster a competitive and dominant tech industry. The proposed executive order had intended to create a voluntary, pre-release benchmarking infrastructure, rather than imposing stringent government mandates. It would have involved leading AI companies, including OpenAI, Anthropic, Google, and xAI, voluntarily granting federal agencies an early screening window—ranging from 14 to 90 days—to stress-test frontier AI systems for potential national security risks, particularly weaponized cyber capabilities. The cancellation leaves a significant void in federal AI policy, as the government now lacks a unified and coherent safety framework. This comes at a critical juncture, with Silicon Valley preparing for a summer filled with the rollout of complex new AI models. Critics suggest this inconsistent policy approach creates industry ambiguity, especially given the administration's reported opposition to state-level AI safety laws. Instead, the government appears to be leaning towards decentralized, independent testing agreements with tech giants through the Department of Commerce. The push for government action on AI safety had gained momentum following growing concerns within the banking industry and other sectors about AI's rapidly advancing capabilities to identify cybersecurity vulnerabilities. Treasury Secretary Scott Bessent and former Federal Reserve Chair Jerome Powell had previously convened meetings with Wall Street CEOs to discuss the cybersecurity risks posed by advanced AI models. Ultimately, the administration's priority appears to be safeguarding American tech corporations and ensuring their continued leadership in AI development, even if it means deferring a comprehensive federal safety framework for now.