Trump, Tech CEOs Sign Voluntary Accord on Super Intelligence Safety
US President Donald Trump and top tech CEOs signed a voluntary accord on frontier AI safety, outlining four layers of controls and audits.
United States President Donald Trump and the chief executives of several leading technology firms signed a joint agreement on Tuesday aimed at governing the safety of frontier computing models, according to a document released after a White House summit on what the administration has termed "Super Intelligence."
The accord, titled "Joint Commitment on Frontier Responsibilities," is described as morally binding and commits companies to a framework of four layers of controls and audits. It carries the signatures of Trump along with Sundar Pichai of Google, Dario Amodei of Anthropic, Mark Zuckerberg of Meta, Greg Brockman of OpenAI, Elon Musk of xAI, and Jensen Huang of Nvidia.
The declaration states that every company is responsible for developing its technology safely and in a manner that builds public trust, and that firms training and deploying frontier models must have robust processes to address issues quickly.
Under the first layer, companies are to maintain strong internal controls to monitor model capabilities and alignment during training and deployment, covering areas such as cybersecurity, biosecurity, and chemical threats, and to prevent models from accessing technical systems in unintended ways.
The second layer requires an internal team to ensure that controls, monitoring, and detection operate as intended and that any problems are fixed. The third calls for an independent external auditor or evaluator to assess whether those controls are working. The fourth mandates an independent board committee to oversee the process, receive reports from internal teams and external auditors, and ensure identified issues are remediated.
The signatories said these layers together should give companies, customers, and the public confidence that the technology is operating as intended. They also agreed to meet regularly to update safety standards and best practices as model capabilities advance.
While the commitment is voluntary, the document acknowledges that the measures could eventually be codified into laws or regulations. It states that implementing the controls and audits is critical to a safe future, regardless of whether they become mandatory.