Zuckerberg Calls For More AI Oversight In New Manifesto

Mark Zuckerberg on Monday published a lengthy letter outlining his thinking on the future of AI, arguing for a broader strategy focused on “individual empowerment” while calling for further regulation by the U.S. government as AI becomes increasingly ingrained in society.

Zuckerberg, in a 6,500-word letter titled “The Future is for Everyone: The Path to a Positive AI Future,” wrote that Meta will resume releasing open-source AI models, which he referred to as a “positive and important force for empowering people.”

AI users should be responsible for how the technology takes form, Zuckerberg wrote, arguing that if “superintelligence is held by a small number of individuals, businesses, government, or AI itself,” it would “naturally lead to outcomes that are less favorable for everyone else.”

He argued that distillation, or the process of creating a new, smaller AI model by transferring knowledge from a more powerful model, should be viewed more favorably despite pushback from the Trump administration: “Some have tried to frame distillation as harmful, but I think it is important to protect the principle that you can learn from anything that you observe.”

Meta will launch a $1 billion “Future Is For Everyone Fund” to support communities near the firm’s data centers, Zuckerberg said, adding that the facilities are “designed to be among the most water-efficient in the world.”

The government should be involved earlier in the regulatory process for new AI models, according to Zuckerberg, who previously said the Trump administration’s proposed 30-day review period was “quite a meaningful amount of time,” arguing there should instead be a “close proactive collaboration” between AI labs and the government rather than a “rigid process and review timeline.”

Meta will also adopt new oversight of model releases that allows its board to say whether models adhere to safety guidelines it approves, Zuckerberg wrote, calling on other companies to adopt similar structures as an “industrywide version.”

“Most other [AI] labs are focused on building AI for companies, governments, or other institutions, so if those labs lead, then the balance of power will favor larger institutions over individuals,” Zuckerberg wrote.

$145 billion. That’s Meta’s projected spending for its current fiscal year as the firm and its mega-cap competitors ramp up a buildout of AI infrastructure, with plans for the Facebook parent to spend $600 billion by 2028.

Zuckerberg has been bullish on the future of AI, and he has been strongly opposed to the technology becoming primarily controlled by a single company, government or small group of firms. In an op-ed for the Wall Street Journal last month, Zuckerberg said AI could soon use the technology to “create and discover extraordinary new things, build new businesses, express our ideas, learn new concepts and improve our lives, health, relationships and careers.” He also said that “putting power in people’s hands” is how “humanity has made the most progress.” Zuckerberg’s approval of distillation comes as the Trump administration has said it would look into whether leading Chinese AI models had been distilled from American models. The U.S. has separately alleged that Chinese firms have smuggled banned AI chips into China from the U.S. to develop their competing models.

Zuckerberg, who co-founded Facebook in 2004 before the firm rebranded as Meta in 2021, has an estimated net worth of $206.2 billion as of Monday.