Meta pushes global AI vision amid race with OpenAI, Anthropic and China

Meta CEO Mark Zuckerberg outlined a vision for the United States to lead the global AI race against competition from China and other tech giants such as OpenAI and Anthropic. In a lengthy essay published on AI's future, Zuckerberg called for the US to focus on open weight models, allowing users to access the rules governing AI behavior without exposing all underlying code.

This approach contrasts with closed models used by competitors like OpenAI and Anthropic, whose AI systems are known as "black boxes" that users cannot inspect. Meta has invested heavily in its AI division, Meta Superintelligence Labs, pouring billions of dollars into developing AI technology. Yesterday, Meta unveiled a new open weight model called Glimmer, derived partly from a closed-source model it announced in April, to provide more powerful AI agents while using less computing power.

The push for open models comes amid growing debate over AI regulation, with President Trump setting up a voluntary review process for advanced AI models. However, Zuckerberg argues for flexible, proactive collaboration with the government instead of a rigid review timeline. Meta also announced a $1 billion fund to support communities around its data centers, as local opposition to data center development grows.