Meta CEO Mark Zuckerberg has outlined his vision for “personal superintelligence,” signaling a potential shift in the company’s AI model release strategy. Zuckerberg’s concept envisions AI assisting individuals in achieving their personal goals.
In a statement, Zuckerberg emphasized that while the benefits of superintelligence should be widely shared, the novel safety concerns associated with such advanced AI necessitate careful consideration regarding what to open source. This marks a potential departure from Meta’s previous approach, where the company’s Llama models were positioned as a key differentiator through open sourcing, aiming to surpass closed models from rivals like OpenAI, xAI, and Google DeepMind.
Zuckerberg had previously acknowledged the possibility of limiting open sourcing if it was deemed irresponsible. While some argue that Llama doesn’t strictly adhere to the definition of open source AI due to Meta’s withholding of its massive training datasets, Zuckerberg’s recent remarks suggest a change in priorities, with open source no longer being the default for Meta’s most cutting-edge AI.
Meta’s competitors typically keep their models closed to maintain control over monetization. Zuckerberg previously noted that Meta’s business model, primarily reliant on internet advertising, doesn’t depend on selling access to AI models. Therefore, “releasing Llama doesn’t undercut our revenue, sustainability, or ability to invest in research like it does for closed providers.”
However, this viewpoint shifted as Meta executives became increasingly focused on surpassing OpenAI’s GPT-4 model during the development of Llama 3. In June 2025, Meta initiated a public AGI sprint, investing $14.3 billion in Scale AI, acquiring Scale’s founder and CEO, and restructuring its AI efforts under Meta Superintelligence Labs. The company has invested heavily in acquiring AI talent and building new data centers.
Recent reports indicate that Meta paused testing on its latest Llama model, Behemoth, to concentrate on developing a closed model. Zuckerberg’s mission to introduce “personal superintelligence” suggests that Meta plans to deliver this technology through its own products, such as augmented reality glasses and virtual reality headsets.
Zuckerberg envisions personal devices like glasses that understand our context by seeing and hearing what we do, interacting with us throughout the day, and becoming our primary computing devices. When questioned about the possibility of Meta keeping its most advanced models closed, a company spokesperson affirmed Meta’s commitment to open source AI while also acknowledging the expectation of training closed source models in the future. “Our position on open source AI is unchanged,” the spokesperson stated. “We plan to continue releasing leading open source models. We haven’t released everything we’ve developed historically and we expect to continue training a mix of open and closed models going forward.”
In 2024, Zuckerberg stated, “Starting next year, we expect future Llama models to become the most advanced in the industry.”
