Apple Bets On A Microsoft AI Veteran And Google’s Gemini To Rescue Siri

Apple has quietly been reshaping its artificial intelligence strategy for months, but the latest move is impossible to ignore. After years of relying on veteran AI chief John Giannandrea to steer everything from Siri to on-device intelligence, the company is pushing through its most dramatic leadership change yet: Giannandrea is leaving, and Apple is handing the keys to its AI future to Amar Subramanya, a senior AI figure coming over from Microsoft.

Subramanya will join Apple as a vice president overseeing AI initiatives under software engineering boss Craig Federighi, a sign that AI is no longer a side project but a core layer of the operating systems that power the iPhone, iPad and Mac.

In an internal note, Tim Cook praised Federighi’s role in driving Apple’s AI efforts so far and said that, with Amar on board, the company is doubling down on the work to deliver a more personal, context-aware version of Siri to users next year. The message is clear: if Apple once appeared cautious in the AI race, it now wants to look decisive.

A leadership shakeup with high stakes

John Giannandrea arrived at Apple with deep search and machine-learning credentials and spent years trying to drag Siri and Apple’s broader AI stack into the modern era. His departure does not just mark the exit of a single executive; it signals that Apple is ready to admit that its first chapter of AI underperformed compared with rivals pushing aggressive chatbots, copilots and generative models. Bringing in a Microsoft AI researcher is also symbolically powerful, underscoring how small the world of top-tier AI talent has become and how aggressively companies are raiding one another’s benches.

For Apple, Subramanya’s job will not simply be about shipping one high-profile feature. He inherits a fragmented AI landscape inside the company, from on-device models that power features like photo search and autocorrect to server-side systems that drive Siri responses. Aligning those efforts under a single strategy while keeping Apple’s strict privacy promises intact may prove to be one of the toughest leadership challenges in the company today.

Siri’s future is tied to Google’s Gemini

At the same time that Apple is reshuffling leadership, it is quietly making a massive bet on outside AI infrastructure. According to industry reports, Apple plans to base its revamped Siri experience on a custom-tuned version of Google’s Gemini model boasting around 1.2 trillion parameters. That is a staggering leap from the roughly 1.5 billion-parameter, Apple-built model currently used for cloud-based Siri processing, and it illustrates just how far behind the company believes its own large-scale models are for certain tasks.

The financial commitment is equally eye-catching. Apple is expected to pay Google on the order of 1 billion dollars a year for access to Gemini and related proprietary technology. That figure simply becomes another line item in a relationship that is already one of the biggest money flows in consumer tech: Google is believed to send Apple around 20 billion dollars annually to remain the default search engine in Safari and other Apple services. In effect, the two rivals are now bound together not only by web search, but by the next generation of AI assistance as well.

A talent drain toward Jony Ive and OpenAI

While Apple is licensing external AI brains, it is simultaneously watching some of its own human brains walk out the door. The company’s famed hardware and interface design organization has been losing veterans to Jony Ive’s design studio, known simply as io, which was recently snapped up by OpenAI as it chases an ambitious "iPhone killer" concept. The vision reportedly centers on a screenless, pocket-sized object that acts as a physical companion to powerful cloud AI, rethinking what a personal device looks like in the age of assistants instead of apps.

According to reporting on the exodus, roughly 40 Apple engineers have jumped to OpenAI in just the past few weeks. Among them are manufacturing design specialist Matt Theobald and human interface lead Cyrus Daniel Irani, both deeply involved in the way Apple products feel, look and get built at scale. Another rising star, Abidur Chowdhury, credited internally with work on the iPhone Air, has also left Apple to join an undisclosed AI startup. For a company that has long prided itself on a cohesive hardware-software design culture, watching that much institutional knowledge exit at once is no small concern.

Can Apple still define the future of devices?

Put together, these shifts paint a revealing picture of where Apple stands in late-stage smartphone maturity and early-stage AI disruption. On one hand, the company is willing to rent Google’s most advanced models and recruit senior leaders from a direct competitor to accelerate progress. On the other, it is being forced to contend with the gravitational pull of OpenAI and Jony Ive’s new hardware experiments, which promise a reimagining of the device paradigm Apple itself helped create with the original iPhone.

The next few years will show whether Amar Subramanya and Craig Federighi can turn Apple’s sprawling AI projects into a coherent, privacy-respecting assistant that feels as polished and indispensable as the best of the company’s hardware. A truly personalized Siri, powered behind the scenes by a gigantic Gemini-based model yet dressed in Apple’s familiar design language, could reassure customers that the company is still capable of leading rather than following. If it stumbles, though, the narrative that Apple is paying rivals for intelligence while losing its own may become difficult to shake.