CEO Mark Zuckerberg didn’t come to Meta’s earnings call on Wednesday promising a breakthrough AI model that would silence the skeptics.
Instead, he offered something more careful: a bet on momentum.
“I expect our first models will be good,” Zuckerberg said in his opening remarks, “but more importantly, we’ll show the rapid trajectory that we’re on, and then I expect us to steadily push the frontier over the course of the year as we continue to release new models.”
In June, Meta launched a new AI initiative, Meta Superintelligence Labs (MSL), headed by former Scale AI CEO Alexandr Wang. When Wall Street sought signs that this big, costly AI reset is working, Zuckerberg had a different ask: patience and faith that a steady drumbeat of releases in 2026 will matter more than a single big reveal.
“The AI strategy articulated on the call may leave some wanting more,” wrote Barclays analyst Ross Sandler in a note to clients, “but there was an underlying confidence and clearly a lot of new things in the works.”
Zuckerberg said he could not yet share details of the company’s AI strategy on the call and that it would roll out its initial set of models and products over the coming months.
“I think my answers to a lot of your questions on this particular call may be somewhat unfulfilling because we’re in this interesting period where we’ve been rebuilding our AI effort. And we’re six months into that, and I’m happy with how it’s going,” he said.
That didn’t stop some analysts from pressing the CEO. When JPMorgan analyst Doug Anmuth asked Zuckerberg for more details on MSL’s progress, Zuckerberg said he had nothing else to add, though he was “very pleased with the quality of the team.”
“The first set of things that we put out, I think, are going to be more about showing the trajectory that we’re on rather than being a single moment in time,” Zuckerberg said.
‘No real substance’ to CEO Mark Zuckerberg’s comments
Brian Mulberry, an analyst at Zacks Investment Management, told Business Insider that there was “no real substance to any of Zuckerberg’s comments that would move our analysis one way or the other.”
“We want to see real bottom-line profits driven by AI, and it seems that Meta is still far from that reality,” Mulberry said.
Roger Beharry Lall, a research director at IDC, told Business Insider that Zuckerberg’s remarks about the company’s coming AI models and their “trajectory” show that the company is ambitious, though the lack of concrete information means its goals “remain largely aspirational.”
Meta can spend big on AI because of ads
Meta’s financials show that its AI work is already improving its advertising business. The company said improvements to how it ranks and shows ads led to 3.5% more clicks on Facebook and over 1% gain in conversions on Instagram in the final quarter of 2025.
Meta’s revenue jumped 24% to $59.9 billion in the last three months of 2025, and the company generated $43.6 billion in free cash flow in 2025 even as it spent heavily on AI infrastructure. Ad impressions jumped 18% from this time last year, and what advertisers paid for each ad rose 6%.
That combination explains why Meta can afford to keep spending on AI even while Zuckerberg asks investors to wait for the bigger breakthroughs.
Meta’s patient approach carries some risk given the company’s recent track record. In August, Business Insider reported that the company was pushing to release the next version of its Llama AI model by the end of the year, which didn’t happen.
Llama’s previous release in April disappointed developers who said it lagged in coding and reasoning, precisely the capabilities that matter most in the AI race that Meta wants to catch up in.
Meta’s new model, called “Mango”, will focus on images and videos, while another one called “Avocado” will be better at coding, The Wall Street Journal reported.
Some analysts said that Zuckerberg’s restraint may simply reflect Meta’s early stage in its AI reset.
“Training the model with this new team is going to take a while,” Mandeep Singh, Bloomberg Intelligence’s global head of technology research, told Business Insider.
To make up ground, he expects Meta to sharpen its focus and specialize in certain areas, rather than “trying to beat everyone everywhere all at once.”
Singh said that Meta’s strong advertising business gave it a lot of runway.
“This kind of growth rate allows you a lot of cushion in terms of taking your time and getting AI,” he said.
Have a tip? Contact Pranav Dixit via email at pranavdixit@protonmail.com or Signal at 1-408-905-9124. Use a personal email address and a nonwork device; here’s our guide to sharing information securely.
