Trump administration gets Big Tech to sign weak, non-binding, AI regulations

Six major AI companies, including Meta, Google, OpenAI, Anthropic, Nvidia, and XAI, have agreed to a non-binding and vague AI regulation known as the "White House Accord on Superintelligence." The document, which appeared on President Trump's Truth Social account, does not require the companies to do anything specific, but it allows them to set their own rules for ensuring safe AI development.

The accord, subtitled the "Joint Commitment on Frontier Responsibilities," emphasizes that each company is responsible for developing its technology safely and building trust with customers and the public. It outlines four layers of controls and audits that signatories should implement, such as robust internal controls to monitor model capabilities and alignment, empowering an internal team to oversee these controls, partnering with an independent external auditor, and having an independent committee of the board to oversee the situation.

However, the document lacks clear definitions of "robust" internal controls, the frequency of external audits, and the evaluation schedule. It states that the participating companies will meet regularly to establish standards and best practices for improving system safety. The accord suggests that it might be beneficial to codify these steps into laws or regulations in the future.

The accord reflects Trump's belief that the AI industry does not need strong regulation, except for the name. On Tuesday, Trump signed an executive order requiring federal government agencies to use the term "Super Intelligence" and the acronym "SI" instead of "Artificial Intelligence" and "AI" in official communications. Trump has defined frontier AI systems as those that go beyond imitation or automation of human intelligence and instead amplify creativity and unlock new forms of creativity.