Zuckerberg Criticizes AI Centralization, Opposes US Ban on Chinese Models

Meta CEO Mark Zuckerberg has voiced strong opposition to the centralization of artificial intelligence (AI) development and criticized potential U.S. bans on Chinese AI models. In an op-ed published in The Wall Street Journal, Zuckerberg advocated for open access to AI, stating it is key to unlocking prosperity and pushing back against what he described as a prevalent narrative of AI-driven doom.

Zuckerberg expressed surprise at the amount of discourse surrounding AI that focuses on apocalyptic scenarios, arguing that humans are fortunate to be living through a significant technological advancement. He directly challenged the views of some executives who predict AI will eliminate most jobs and diminish human relevance, questioning the rush to build such a future. "The notion that AI is so dangerous that the only safe path is an extreme concentration of power seems dangerous," Zuckerberg wrote. "Historically, hoping that an absolute power will benevolently provide for humanity if sufficiently enlightened hasn’t led to safe or positive outcomes."

His comments come as the Trump administration is reportedly preparing an executive order that would require tech giants to voluntarily submit new AI models for a 30-day review period, a move prompted by cybersecurity concerns following incidents involving OpenAI and Anthropic. Zuckerberg acknowledged that new issues related to AI require consideration but maintained that distributing the technology creates significantly more value than harm. He noted that even a 30- or 60-day review period could be substantial given the rapid pace of AI development.

U.S. officials are also reportedly considering restrictions on American companies accessing Chinese AI models, particularly after accusations that China's Moonshot AI may have used stolen data from Anthropic for its Kimi K3 release. The Federal Communications Commission (FCC) recently banned Chinese humanoid robots, citing national security risks and the potential for surveillance or data collection by malicious actors. However, Zuckerberg doubted the efficacy of banning open-source models from different countries, suggesting that policymakers should instead prioritize accelerating domestic AI advancements.

This stance aligns with Meta's significant investment in its AI capabilities, referred to as "superintelligence" efforts, which includes substantial spending on memory chips and data centers. Despite these investments, Meta currently trails behind competitors like OpenAI and Anthropic in areas such as chatbot performance. Federal restrictions on launching new AI models could further impact Meta's competitive position.

Zuckerberg also addressed concerns about AI's impact on employment, stating that the development of AI infrastructure has led to job creation. He believes that widespread AI accessibility will ultimately lead to more jobs, particularly for entrepreneurs starting small businesses, rather than fewer. Data from Challenger, Gray & Christmas indicates that AI has been cited in a significant number of job cut announcements since 2023, though the overall labor market has shown resilience. Zuckerberg reiterated that the concern about job displacement has not materialized as feared, particularly as AI spurs demand for infrastructure development and related training programs.