Meta's Superintelligence Labs Delivers First AI Models After Six Months
Meta's Superintelligence Labs delivered its first AI models internally after six months of development, CTO Andrew Bosworth said at Davos. This milestone follows codenamed Avocado and Mango projects and signals progress in Meta's AI strategy that could strengthen ad ranking and user engagement tools.
1. Jefferies Analysts Highlight Valuation Upside After Recent Pullback
Meta shares jumped over 5% following a bullish note from Jefferies, which reiterated its Buy rating and set a $910 price target, implying roughly 44% upside from recent levels. The stock has traded about 18% below its post–earnings high since the third‐quarter report, when concerns over capital spending on AI prompted profit taking. Jefferies pointed to Meta’s strong ad revenue growth—26% year-over-year in Q3—driven by enhancements to its AI ranking systems, and noted free cash flow of nearly $11 billion in the quarter, as key underpinnings for the valuation case.
2. Meta to Monetize Threads with Global Ad Rollout
Building on its rapid user growth, Meta will begin serving advertisements on its Threads platform to all users worldwide next week. Threads, launched in July 2023, now boasts over 400 million monthly active users and has surpassed its main competitor in daily mobile sessions. The rollout follows successful tests in select markets and is expected to add a new revenue stream atop Meta’s AI-powered ad tools, which already generate more than $60 billion in annualized revenue run rate.
3. In-House AI Lab Delivers First High-Profile Models
Meta’s newly formed Superintelligence Labs team has produced its first proprietary AI models internally, according to CTO Andrew Bosworth at the World Economic Forum in Davos. Six months into operation, the lab’s text and vision models have demonstrated “very good” performance in preliminary tests. These efforts complement Meta’s existing Llama 4 family of large language models and signal the company’s ramped-up investment in building core AI capabilities for future consumer and enterprise offerings.
4. Meta Executive Criticizes EU Rules for Slowing AI Deployment
Nicola Mendelsohn, head of Meta’s global business group, warned on Bloomberg Television that the current European regulatory framework is making it more difficult for companies to roll out and adopt new AI products. Speaking from Davos, she said that business leaders across industries are increasingly frustrated by protracted approval processes and inconsistent standards, which risk putting European firms at a competitive disadvantage in the global AI race.