Meta CEO Mark Zuckerberg has issued a strong defence of artificial intelligence, arguing that concerns about the technology are often overstated and warning that the biggest risk is allowing a single government or entity to gain too much control over advanced AI.


In a 6,500-word manifesto released on August 10, Zuckerberg outlined his vision of AI as a tool for expanding individual capabilities and economic opportunity. He said the technology should be widely accessible, particularly in democratic countries, while humanity retains control over increasingly powerful systems, Axios reports


"One of the ways to achieve both individual empowerment and checks and balances is to make sure you put the technology in everybody's hands," Zuckerberg told Axios ahead of the manifesto's release.


"I'm personally more worried about centralisation than I am about any of the specific risks others are talking about," he said.


Zuckerberg argued that excessive regulation in the United States could weaken the country's position in the global AI race, particularly as China and other countries seek to narrow the technological gap.


"Any policy that slows American model releases — even by a month — could add significant risk to American leadership while letting foreign models race ahead," he wrote.


He said Meta's goal is for everyone to have free or affordable access to AI tools, with paid services using a "dynamic auction mechanism" to obtain computing resources at the lowest possible price.


"We believe this will be true with AI as well, and the abundance of the future can be shared by everyone," he wrote. "We also believe that the values that got us to this point — like liberty, open inquiry, free enterprise, and equal opportunity — are also the right values to build a positive future."


Zuckerberg also said Meta plans to resume releasing some open-source AI models. The company said it would open the weights of its Muse Glimmer model and plans to release the weights of a version of its Muse Spark 1.2 foundational model in the coming weeks.


He said Meta's independent board would have authority to approve safety criteria for model releases and review whether those standards had been met.


Zuckerberg also outlined plans to address concerns surrounding Meta's expanding data-centre footprint. The company is launching the Future is for Everyone Fund with $1 billion for U.S. communities where Meta operates data centres. The initiative will support teachers, first responders, and energy and water infrastructure.


On privacy, Zuckerberg said future Meta AI agents would offer stronger protections than the company's current AI services.


"We believe that people should have a fully private mode for personal agents where even Meta or any other service provider cannot see or grant access to your information."


By Sabina Mammadli