Apple Newsroom:

Apple today announced John Giannandrea, Apple’s senior vice president for Machine Learning and AI Strategy, is stepping down from his position and will serve as an advisor to the company before retiring in the spring of 2026. Apple also announced that renowned AI researcher Amar Subramanya has joined Apple as vice president of AI, reporting to Craig Federighi. Subramanya will be leading critical areas, including Apple Foundation Models, ML research, and AI Safety and Evaluation. The balance of Giannandrea’s organization will shift to Sabih Khan and Eddy Cue to align closer with similar organizations.

JG’s departure comes as a surprise to absolutely no one after he was effectively sidelined by Tim Cook following the delay of the “more personalized Siri” features, which relied heavily on Apple Intelligence—the rollout of which has been less than stellar. Apple is positioning this as his “retiring,” but JG has long been a man without a portfolio. I’m surprised only that it’s taken this long.

JG was seemingly too focused on research and development and not enough on shipping products (in Apple terms, he was perhaps good at his “Category 1”—AI research—and not so great at his “Category 3”—making that work available for others to successfully perform their Category 1).

Breaking up JG’s organization makes sense, then. (My understanding is it was a mess—apparently the admin he brought over with him from Google was running the team.) Subramanya keeps JG’s research and foundational AI portfolio (under Federighi’s SWE—Software Engineering), while I’ll guess that AI infrastructure, which wouldn’t fit well under SWE, shifts to Khan (Apple’s COO), and front-end and related services lands with Cue, who owns Services (like the App Store and App Store Connect). Fortunately, Cue’s and Federighi’s teams have a lot of experience working together to deliver products (Xcode Cloud or In App Purchase are but two examples), so I’m confident this bodes well for the future of Apple Intelligence.

I don’t know Subramanya (I suppose he’s “renowned” among AI researchers?). He left Microsoft after only five months to join Apple, after spending 16 years at Google (twice that of Giannandrea). Is another long-time Google executive the right move here? If he’s willing to leave Microsoft after just five months, what happens when Meta or OpenAI come calling with a yet-better offer? On the other hand, if Apple does partner with Google to use Gemini, having someone with deep familiarity of that team could prove valuable. I’m cautiously optimistic here.

(Amusingly, Subramanya’s LinkedIn still shows his last post, where he is “super excited” to start at Microsoft and was “feeling deeply energized” a week in, describing Microsoft’s culture as “refreshingly low ego yet bursting with ambition.” Either something changed or Apple made a really great offer.)