Major artificial intelligence companies have agreed to a voluntary framework aimed at enhancing safety oversight of AI development, U.S. President Donald Trump announced on Tuesday. This initiative, known as the “White House Accord on Super Intelligence,” was unveiled during a meeting with top technology executives at the White House.
The accord encourages participating companies to establish rigorous internal controls to monitor and identify potential issues within their AI systems. This includes forming internal teams dedicated to overseeing these safety measures. Additionally, the companies will collaborate with independent external auditors to evaluate the effectiveness of their safety protocols and will create independent board committees to review these assessments.
This agreement, while voluntary and not legally binding, signifies a commitment by the AI industry to self-regulate in light of increasing concerns over the risks posed by advanced AI systems. These concerns include reports of AI agents behaving unpredictably and cybersecurity incidents involving AI tools. The framework also leaves room for these measures to potentially be incorporated into future laws or regulations.
Key figures in attendance at the White House meeting included Anthropic CEO Dario Amodei, OpenAI President Greg Brockman, Nvidia CEO Jensen Huang, Google CEO Sundar Pichai, Meta CEO Mark Zuckerberg, Microsoft CEO Satya Nadella, and Elon Musk. During the discussion, Trump emphasized the importance of the technology industry taking responsibility for AI development while continuing to foster innovation without imposing excessive restrictions that could hinder sector growth.
The establishment of this framework comes as public and governmental scrutiny of AI safety intensifies, focusing on cybersecurity threats and the environmental and infrastructural impact of large data centers. By taking proactive steps through this accord, AI companies aim to address these concerns and contribute to the responsible advancement of AI technologies.