<p>President Donald Trump on September 29, 2026, released an agreement with technology CEOs to self-regulate artificial intelligence, referred to by Trump as "super intelligence," amid growing public concerns about the technology's development.</p>
<p>The agreement, titled the White House Accord on Super Intelligence Joint Commitment on Frontier Responsibilities, outlines guidelines for companies developing AI frontier models to implement internal controls, establish an internal oversight team, conduct external audits, and create an independent oversight board.</p>
<p>Trump announced the agreement following a meeting with leading AI companies and Cabinet officials. Signatories include Google CEO Sundar Pichai, Meta CEO Mark Zuckerberg, Anthropic CEO Dario Amodei, SpaceXAI CEO Elon Musk, Nvidia CEO Jensen Huang, and OpenAI co-founder and President Greg Brockman. While Zuckerberg and Huang have expressed opposition to such regulations, the other CEOs have shown support.</p>
<p>The accord emphasizes the responsibility of companies to develop technology safely and build trust with customers and the public. It specifies that companies training and deploying frontier models should have robust internal processes to ensure their technology functions as intended and that any issues are promptly addressed. The agreement includes four key layers of controls and audits:</p>
<ol> <li>Implement robust internal controls to monitor the capabilities and alignment of models during training and deployment, focusing on areas such as cybersecurity, biosecurity, and chemical threats, to prevent unintended access to technical systems.</li> <li>Establish an internal team to ensure that all controls, monitoring, and detection are functioning correctly and that any issues are resolved.</li> <li>Engage an independent external auditor to conduct assessments of the operational effectiveness of the controls, monitoring, and detection systems.</li> <li>Create an independent committee within the board of directors to oversee the operations of the controls and auditors, ensuring that identified issues are addressed.</li> </ol>
<p>These measures aim to provide confidence to companies, customers, and the public regarding the safe operation of AI technology.</p>
<p>The participating companies will convene regularly to establish standards and best practices for enhancing the safety of their systems.</p>
<p>Over time, the agreement suggests that it may be appropriate to codify these measures into laws or regulations. Regardless of legal requirements, the companies involved are committed to implementing these controls and audits to ensure a secure future for all.</p>