Meta building an AI ‘Zuckerberg’ for staff

Meta is reportedly building an internal artificial intelligence version of Mark Zuckerberg so employees can ask “the boss” questions without waiting for the real one. (ft.com) (theguardian.com) The Financial Times reported on April 12 that the project is a photorealistic three-dimensional character trained on Zuckerberg’s voice, words, mannerisms and recent thinking on company strategy. The report said Zuckerberg is helping train and test it himself. (ft.com) The reported goal inside Meta is scale: one chief executive cannot answer tens of thousands of employees one by one, but a chatbot or avatar can reply around the clock in his style. The Guardian said the system is being pitched as a way for workers to “talk to the boss.” (theguardian.com) (ft.com) This fits Meta’s broader push to put artificial intelligence into more parts of the company at the same time it reshapes itself around the technology. On April 8, Meta announced Muse Spark, a new model from Meta Superintelligence Labs that it said will power the Meta AI app, website, WhatsApp, Instagram, Facebook, Messenger and its artificial intelligence glasses. (about.fb.com) Meta has also been using artificial intelligence as a management tool outside consumer products. In January, the company said it was expanding a Meta AI business assistant for advertisers that can remember goals and offer personalized recommendations. (about.fb.com) The internal Zuckerberg character lands after a year in which Meta tightened performance management while spending heavily on artificial intelligence. Reports in February 2025 said Meta planned to cut about 5% of its workforce in performance-based layoffs, while other reports said the company told staff in January 2026 that capital spending this year could reach $115 billion to $135 billion. (finance.yahoo.com) (moneycontrol.com) The idea is not just a text bot with canned answers. The Financial Times said Meta is pursuing photorealistic characters as part of Zuckerberg’s push for what he calls “personal superintelligence,” a phrase the company used again in its December 2025 year-end summary. (ft.com) (about.fb.com) That raises a practical question inside any large company: whether employees are getting the chief executive’s actual judgment, a synthesis of his past statements, or a polished approximation designed for consistency. Meta had not publicly announced the tool as of April 14, and the reported details so far come from people familiar with the project rather than a formal company launch. (ft.com) (theguardian.com) If Meta deploys it widely, the company would be turning the chief executive into a permanent internal interface: always available, on-message and machine-generated. For a company already putting artificial intelligence into ads, apps and glasses, this would extend that logic to the corner office. (about.fb.com 1) (about.fb.com 2)