United States President Donald Trump met with artificial intelligence executives at the White House to establish a voluntary agreement called the White House Accord on Super Intelligence. The pact, signed by executives from Meta, Nvidia, Google, OpenAI, xAI, and Anthropic on September 30, 2026, encourages companies to implement internal security measures with four layers of control. These include maintaining internal controls, establishing an internal oversight team, collaborating with independent auditors, and creating an independent board committee for review.
The agreement is voluntary and does not impose penalties for non-compliance, nor does it require public disclosure of audit results or government enforcement. However, it allows for the possibility that these measures could become mandatory in the future. Trump stated that there would be a significant self-policing aspect to the agreement, describing it as “morally binding.”
Google CEO Sundar Pichai supported the accord, emphasizing the commitment to establish norms and build public confidence in AI development. The agreement comes despite previous calls from some executives for government-mandated safety standards.
In September, OpenAI's Chris Lehane advocated for Congress to create mandatory national AI safety standards, while Anthropic's Dario Amodei and Google DeepMind's Demis Hassabis called for more stringent regulations and oversight. Concerns have been raised internally at these companies regarding the monitoring of advanced AI systems.
Legislative efforts for binding regulations are being led by Congress, with proposals such as the Ban Artificial Superintelligence Act introduced by Senator Bernie Sanders and Representative Greg Casar, which aims to ban the development of artificial superintelligence and pause advanced AI development until safety rules are established.
Additionally, the AI Systems Transparency Act, introduced by a bipartisan group of senators, would require AI companies to disclose data collection practices and safety measures. Trump has also directed agencies to use the term “Super Intelligence” in place of “artificial intelligence” in official communications, outlining a new executive order to define this term within 60 days.