Ted Cruz AI bill could let firms bribe Trump to avoid safety laws, critics warn (arstechnica.com)

Senator Ted Cruz has introduced the SANDBOX Act, a controversial AI policy framework that would allow Big Tech firms to sidestep federal laws restricting AI experiments for up to a decade. Designed to promote a "light-touch" regulatory approach, the bill aims to advance American leadership in AI by blocking what Cruz terms "burdensome" state and foreign regulations, prioritizing innovation over immediate safeguards. Under the act, companies could apply for temporary exemptions from enforcement of certain federal laws, detailing risks and mitigation strategies, with agencies reviewing these requests—but with the White House Office of Science and Technology Policy (OSTP) holding the ultimate veto power over agency decisions. This concentration of authority in the OSTP has raised alarms among critics who fear it could pave the way for undue influence by AI companies on regulators, potentially allowing risky AI experiments to proceed without adequate oversight. The bill allows for initial two-year moratoriums on enforcement of AI rules, renewable up to five times, cumulatively permitting a decade of minimal regulatory intervention. Further, Cruz’s framework suggests Congress should ultimately enshrine any successful moratoriums as permanent law, effectively enabling long-term deregulation of AI development. For the AI/ML community, this bill signals a major shift toward prioritized innovation and rapid experimentation at the possible expense of public safety and ethical standards. It underscores ongoing tensions between fostering technological advancement and implementing robust governance to address the far-reaching risks of AI systems in society.