Apple's AI Division Faces 'Turmoil'

Apple's AI division is facing a significant leadership reshuffle and strategic pivot after failing to deliver on a revamped Siri. Amar Subramanya, formerly of Google and Microsoft, has been appointed as the new VP of AI, reporting to Craig Federighi. This follows John Giannandrea's planned retirement in Spring 2026, after joining Apple from Google in 2018. The company is reorganizing its AI structure, moving away from a centralized AI team to a functional structure. Parts of Giannandrea's former team are being shifted to Sabih Khan and Eddy Cue, aligning AI functions with operations and services. This restructuring is viewed as a strategic alignment rather than a knee-jerk reaction to recent departures. Apple's struggles in the AI race have led to internal frustration and talent exodus, with some AI researchers joining competitors like Google DeepMind and Meta. The company's decision to potentially outsource some AI tech to Google, specifically using the Gemini model for Siri, has reportedly caused discontent among staff. Apple may pay Google $1 billion annually to license Gemini AI to revamp Siri. The delayed Siri update, originally slated for iOS 26.4, has raised concerns among internal testers regarding its performance. Apple is aiming for on-device processing for privacy, but this has created technical limitations. Despite these challenges, Apple is betting on a local-first intelligence approach with its "Apple Intelligence" to keep user data private. Jony Ive's departure in 2019 also continues to have ripple effects on Apple's design philosophy. Many of Ive's former team members have left, and his design influences are gradually being phased out. While Ive's departure initially sparked concerns, Apple emphasized its continued collaboration with his new company, LoveFrom. Despite the "turmoil," Apple is moving forward with AI initiatives, including a planned AI search tool to compete with OpenAI, potentially integrated with Siri, Safari, and Spotlight. The company has also been testing Google's Gemini AI model for Siri. Apple is preparing two new versions of Siri, with one update tapping into personal data and a more ambitious overhaul built around a chatbot-style interface. Apple's AI strategy emphasizes personal data protection compared to competitors like Google and Microsoft. Apple's approach highlights a broader industry split: cloud-centric AI firms are racing to scale huge models, while Apple is betting on on-device inference, efficiency and privacy. However, Apple researchers have found that AI reasoning models face a "complete accuracy collapse" when presented with highly complex problems. Looking ahead to 2026, Apple has several major product launches planned, including AI glasses, a smart-home-control tablet, and a foldable iPhone. Successful AI integration will be crucial to driving hardware upgrades, as Apple is unlikely to charge for its AI features.