AI firms to establish independent committees for safety oversight

Google, Anthropic, Meta, OpenAI, xAI and Nvidia have agreed to establish independent committees to oversee the development and safety of artificial intelligence systems, according to a document signed by US President Donald Trump and the companies' executives, Report informs.

Trump shared the document on his Truth Social account.

The document outlines a four-level oversight framework for AI development. Under the proposed framework, the companies would establish independent committees within or alongside their boards of directors to oversee internal and external auditors, as well as teams responsible for AI safety.

The companies also agreed to involve external auditors and experts to independently assess existing oversight mechanisms. In addition, each organization would operate dedicated systems to monitor the capabilities of AI models during training and deployment and assess their compliance with safety requirements.

Particular attention would be given to cybersecurity and potential biological and chemical threats.

The companies would also establish specialized units to monitor the operation of oversight and monitoring systems and help identify potential threats.

According to the document, the measures could eventually be incorporated into legislation. The companies also agreed to hold regular consultations to develop common standards and best practices for AI safety.