
In a fresh internal memo (shoutout to Axios for the scoop), Meta’s chief product officer, Chris Cox, laid out the blueprint for a reorg that splits its AI ops into two turbocharged teams: the AI Products crew and the AGI Foundations squad. This move isn’t just about changing titles—it’s Meta’s latest attempt to flex smarter, build faster, and stay relevant in the generative AI race.
So what’s the deal?
Team #1: AI Products—led by Connor Hayes—is now steering the ship on everything customer-facing. That includes the Meta AI assistant (yes, the chatbot trying to be your new BFF), Meta’s AI Studio, and the AI baked into Facebook, Instagram, and WhatsApp.
Team #2: AGI Foundations—co-led by Ahmad Al-Dahle and Amir Frenkel—is where the deep tech magic happens. Think bigger-picture research, smarter Llama models, and next-gen capabilities in reasoning, multimedia, and voice. In other words, these folks are working on the stuff that’s trying to compete with OpenAI’s GPT-niverse.
Interestingly, FAIR (Meta’s OG research group) is staying independent for now, though one multimedia team is hopping over to AGI Foundations. Also worth noting: No layoffs, no execs jumping ship (well, aside from a few heading to French rival Mistral), but yes—leaders are being reshuffled to optimize the build machine.
Why the split? According to Cox, it’s all about “ownership” and speeding up delivery without getting tangled in too many dependencies. (Translation: Move fast, don’t break each other’s code.)
With OpenAI launching 4o and Google demoing next-gen Gemini, Meta knows the pressure’s on. The Llama-for-Startups program and LlamaCon event weren’t just PR—they were signals. Meta wants to be more than a social empire. It’s angling to be a serious AI heavyweight too.