Apple has announced the departure of John Giannandrea, its longtime senior vice president of Machine Learning and AI Strategy, who is stepping down after seven years at the company. Giannandrea, who joined Apple in 2018 from Google—where he led AI and Search—was instrumental in building Apple’s AI and machine learning infrastructure. However, his exit follows growing internal turmoil and setbacks in Apple’s ambitious AI rollout, particularly with the delayed and underperforming upgrade to Siri. The company cited organizational restructuring as the reason for the leadership change. Giannandrea will transition into an advisory role until his retirement in spring 2026. His responsibilities have been redistributed: Craig Federighi, Apple’s software chief, now oversees AI efforts more directly, while Sabih Khan and Eddy Cue will lead key AI and ML teams. Giannandrea’s departure comes amid mounting criticism of Apple’s AI strategy. The launch of Apple Intelligence, introduced in October 2024, was marred by significant flaws. Early versions generated false headlines—such as falsely claiming Luigi Mangione had shot himself and that darts player Luke Littler had won a championship before the final—drawing backlash from media outlets like the BBC. Siri’s promised overhaul, intended to deliver a more personalized and intelligent assistant, was delayed indefinitely after internal testing revealed critical failures. Bloomberg reported that even Craig Federighi struggled to get basic features working during demos. Internal reports suggest deeper problems: poor coordination between AI and marketing teams, budget mismanagement, and a leadership vacuum that led employees to mock Giannandrea’s team as “AI/MLess.” There was also a noticeable exodus of AI talent to rivals like OpenAI, Google, and Meta. To address these challenges, Apple has hired Amar Subramanya as its new vice president of AI. Subramanya brings extensive experience from Google, where he spent 16 years and led engineering for the Gemini Assistant. Most recently, he served as corporate VP of AI at Microsoft. His appointment signals Apple’s intent to accelerate AI innovation and strengthen its technical leadership. Subramanya will report to Federighi and lead Apple Foundation Models, machine learning research, and AI safety. Apple emphasized his expertise in translating AI research into real-world products—a crucial skill as the company races to catch up with competitors. Apple’s AI strategy has long emphasized privacy and on-device processing using Apple Silicon chips. Unlike rivals relying on massive cloud-based models, Apple prioritizes local computation and minimal data collection. However, this approach limits model size and training data, forcing Apple to rely on licensed and synthetic datasets rather than the vast real-world data used by OpenAI and Google. Despite these constraints, Apple is reportedly leaning on Google’s Gemini AI model to power the next version of Siri—a surprising alliance given the long-standing rivalry between the two tech giants. Apple CEO Tim Cook praised Giannandrea’s foundational contributions and welcomed Subramanya’s expertise, reaffirming that AI remains central to Apple’s future. With Subramanya’s arrival and Federighi’s expanded oversight, Apple aims to deliver a more capable and trustworthy Siri in spring 2025, marking a pivotal moment in its AI journey. The company’s ability to balance privacy, performance, and innovation will be critical in determining whether it can regain its position in the AI race.