But, according to Gurman, it didn’t quite work out that way. Apple seemed to lag as rivals pushed ahead with new AI features. “Apple Intelligence” is the most obvious example here with a slow rollout and features getting delayed. Internally, the feeling was that the unified structure wasn’t delivering fast enough.
Apple’s “Personal Siri” experiment didn’t quite work out the way Apple intended. | Image credit — Apple
Hitting the reset button
So, here’s what Apple is doing next: The company is now breaking up that central AI team and moving parts back into its usual structure, where teams are grouped by what they do (software, hardware). Siri development is now back under the main software engineering group led by Craig Federighi. Apple’s robotics explorations are moving over to John Ternus’s hardware engineering team. Gurman suggests this isn’t just house cleaning, but rather that Apple felt the old setup wasn’t delivering the goods.This is all happening while competitors are going full throttle on AI. Google keeps putting its Gemini AI into everything from Search, to Pixel phones, to even Samsung Galaxy phones. Microsoft has its Copilot assistant popping up all over Windows and Office. Meta is pushing AI in its apps and even into its Ray-Ban smart glasses. The pressure is definitely on.
Giannandrea (JG) is still overseeing the core AI technology – the basic models behind Apple Intelligence, research, and data work. The focus seems narrower now, concentrating on the foundations while other teams handle putting it into products. There’s even talk, Gurman notes, that Apple might not replace JG if he leaves, potentially ending this whole “AI chief” experiment.
Now, the question is: is this a smart pivot or a sign that Apple is struggling? Moving teams back to their functional homes feels very Apple – practical and focused. It could definitely streamline getting AI into actual products, but it also makes you wonder if Apple decided its big AI push needed a different strategy. We’ll have to see if this helps Apple catch up, but it’s clear they know something needed to change to better compete in the AI race.
