CEO Mark Zuckerberg says Meta's AI story is about where it's headed, not just one model

During Meta's recent earnings call, CEO Mark Zuckerberg refrained from unveiling a groundbreaking AI model. Instead, he focused on the company's commitment to a gradual development in AI technology. "I expect our first models will be good," Zuckerberg stated, highlighting a forward-looking approach that prioritizes the ongoing evolution of their AI initiatives. In June, Meta launched its ambitious AI project, Meta Superintelligence Labs (MSL), led by former Scale AI CEO Alexandr Wang. While investors looked for immediate signs of success from this substantial investment in AI, Zuckerberg urged them to remain patient, suggesting that the accumulation of releases in 2026 would be more significant than any single announcement. Barclays analyst Ross Sandler noted that while Zuckerberg's presentation might leave some wanting more, it also reflected a solid confidence in the company's future AI developments. However, he acknowledged that the strategy shared during the call could appear vague, as it is still in the early stages. Zuckerberg admitted that he could not provide extensive details about the AI strategy but reassured stakeholders that the company would introduce its initial models and products soon. "My responses might be a bit unfulfilling because we’re in a unique phase of restructuring our AI efforts, and I’m pleased with our progress after six months," he explained. Despite inquiries from analysts, including JPMorgan’s Doug Anmuth, for more specifics about MSL’s advancements, Zuckerberg maintained that he had no further updates, expressing satisfaction with the team’s quality. He reiterated that the upcoming releases would demonstrate the company’s trajectory in AI rather than represent a singular breakthrough. Analysts like Brian Mulberry from Zacks Investment Management expressed skepticism about Zuckerberg's comments, stating there was no substantial information to influence market analysis. He emphasized the need to see tangible profits from AI to gauge Meta’s true progress. Roger Beharry Lall, a research director at IDC, remarked on Zuckerberg's focus on future AI models, suggesting that while the ambitions are clear, the lack of detailed information renders these goals primarily aspirational. Financial reports indicate that Meta’s advancements in AI are already benefiting its advertising operations, with a 3.5% increase in clicks on Facebook and a 1% rise in conversions on Instagram during the last quarter of 2025. The company reported a 24% revenue increase, reaching $59.9 billion, and generated $43.6 billion in free cash flow, even with significant investments in AI infrastructure. Despite the promising financial outlook, Meta’s cautious strategy carries risks, particularly given its recent challenges. An earlier expectation to launch the next version of the Llama AI model by year-end was unmet, and feedback indicated that previous iterations fell short in critical areas. New models like “Mango,” focused on multimedia, and “Avocado,” aimed at improving coding capabilities, are in development. Experts suggest that Zuckerberg's measured approach may stem from the early phase of Meta's AI overhaul. Mandeep Singh from Bloomberg Intelligence noted that training these new models will take time, and he anticipates Meta will concentrate on specific areas to gain an edge in the competitive AI landscape. Singh emphasized that Meta's robust advertising business provides a buffer, allowing the company to invest in AI at a deliberate pace.