Trump signs 'Super Intelligence' safety accord with tech CEOs

US President Donald Trump has signed a joint commitment with leading technology executives to establish safeguards around advanced artificial intelligence systems, posting the "White House Accord on Super Intelligence" to his Truth Social platform as the administration outlined a four-layer framework for internal controls and independent oversight.

Four-layer oversight framework

The agreement, titled "Joint Commitment on Frontier Responsibilities," calls on companies developing frontier AI models to implement robust internal processes ensuring systems operate as intended and potential problems are identified swiftly. According to the document, each participating firm must establish four distinct layers of controls and audits covering both model development and deployment.

The first layer requires internal controls monitoring model capabilities and alignment, specifically addressing risks involving cybersecurity, biosecurity, and chemical threats while preventing systems from hacking or accessing technical infrastructure in unintended ways. A second layer mandates internal teams responsible for verifying that controls, monitoring, and detection systems function properly and that identified problems are addressed immediately.

Under the third layer, companies must engage independent external auditors or evaluators to assess whether their controls and monitoring systems operate as intended, providing verification beyond internal processes. The fourth layer establishes an independent board committee to oversee the entire process, receive reports from internal teams and outside evaluators, and ensure identified issues reach resolution.

Industry commitment and regulation

The measures are intended to give companies, customers, and the public greater confidence that advanced AI systems operate safely, with participating firms committing to meet regularly to establish standards and best practices for system safety. The document acknowledges the possibility of future government regulation, noting the measures could eventually be incorporated into laws or regulations.

Regardless of whether such requirements become mandatory, the signatories said that implementing the controls and audits remains critical to developing the technology safely and committed their companies to adopting the framework. Trump signed the agreement alongside Google CEO Sundar Pichai, Meta CEO Mark Zuckerberg, Nvidia CEO Jensen Huang, OpenAI co-founder Greg Brockman, Anthropic CEO Dario Amodei, and xAI CEO Elon Musk.