Meta moves engineers to AI unit

Meta has started moving senior engineers into a new Applied AI Engineering group as it pushes artificial intelligence tools deeper into product building. (economictimes.indiatimes.com) The Economic Times reported on April 12 that some of those transfers were compulsory, not voluntary, and that the new team is meant to help product groups use artificial intelligence in day-to-day engineering work. (economictimes.indiatimes.com) At the same time, Meta is rolling out a new internal model called Muse Spark, which the company introduced on April 8 as the first model from Meta Superintelligence Labs. Meta said Muse Spark already powers the Meta AI app and website and will expand to WhatsApp, Instagram, Facebook, Messenger and its artificial intelligence glasses in the coming weeks. (about.fb.com) Muse Spark is a multimodal model, which means it can work across text, images and other inputs instead of just words. Meta said the model is built for its own products and is designed to make Meta AI faster while adding features that can cite recommendations and content shared across its apps. (about.fb.com) The reassignments show Meta is putting less distance between research teams and the engineers who ship consumer features. In a January 28 company outlook, Meta said 2026 would be the year artificial intelligence “transforms how we work” across the company, not just inside standalone research groups. (about.fb.com) That push is also tied to a broader reorganization around Meta Superintelligence Labs, the artificial intelligence division led by Alexandr Wang, who joined Meta in June 2025 as the company’s first chief artificial intelligence officer. Meta says Wang now leads the company’s superintelligence effort from strategy through execution. (meta.com) Meta has been building the hardware to support that strategy too. In March, the company said it was developing and deploying four new generations of custom chips within two years to handle recommendation systems and generative artificial intelligence workloads more efficiently. (about.fb.com) Reuters reported on April 9 that investors welcomed Muse Spark as a sign Meta was trying to close gaps with OpenAI, Google and Anthropic; Meta shares were up nearly 7% after the announcement. The same report said Muse Spark came from a costly team Meta assembled last year to speed up its artificial intelligence race. (economictimes.indiatimes.com) Meta has not publicly detailed how many engineers have been reassigned into Applied AI Engineering. But the mix of mandatory transfers, a new product-focused model and fresh infrastructure spending points in one direction: Meta wants artificial intelligence embedded in how its products are built, not parked on the side as a research project. (economictimes.indiatimes.com)