.
Image: Klaudia Radecka/NurPhoto via Getty Images
Apple is making a bold shift in its artificial-intelligence leadership, installing Amar Subramandya, formerly of Microsoft and Google, as its new vice president of AI.
He replaces long-time AI chief John Giannandrea, who will remain at Apple in an advisory capacity until his retirement in spring 2026.
Subramandya will report to software head Craig Federighi and lead Apple’s AI work across “foundation models, machine-learning research, and AI safety and evaluation.”
For years, Apple has been criticised for lagging behind rivals in the race to integrate cutting-edge AI features into flagship products.
While competitors like Samsung Electronics began shipping AI-infused experiences quickly, Apple’s rollout of generative AI, particularly under the umbrella of Apple Intelligence, has felt cautious, incremental, and in many cases delayed.
A prominent example: the long-promised, AI-powered overhaul of voice assistant Siri has been postponed until 2026 as Apple grapples with the complexity of delivering a truly robust, private-first AI assistant.
By bringing in an AI leader who has hands-on experience at both Microsoft and Google, including leading engineering for Google’s generative-AI assistant, Apple appears to be shifting gears from research-first to product-first: a pivot towards embedding powerful AI across its ecosystem in a way that can scale
Faster AI-powered features across devices
With Subramandya overseeing Apple Foundation Models and machine-learning research, customers may see more intelligent, on-device experiences, from smarter Siri to AI-assisted editing on iPhone, iPad, and Mac.
Stronger emphasis on AI safety and evaluation
Given Apple’s public commitment to user privacy and on-device processing, the new structure could help Apple navigate trade-offs between power and privacy.
A more unified AI strategy under software leadership
Having the AI VP report to Apple’s software chief could suggest a tighter integration between AI research and actual product development and perhaps a recognition that AI success hinges less on isolated lab breakthroughs than on seamless, cross-device integration.