Apple Appoints New AI Chief with Google, Microsoft Experience

Apple has appointed a new head of artificial intelligence, bringing in a seasoned executive with deep experience at its primary rivals. Amar Subramanya, a former Microsoft leader who spent 16 years at Google, is taking over the role from John Giannandrea, who is departing the company. This leadership change follows a turbulent period for Apple’s AI initiatives, marked by product delays and public missteps that have left the tech giant playing catch-up in a fiercely competitive field.

Subramanya’s appointment is a strategic move. His recent work leading engineering for the Gemini Assistant at Google provides him with an intimate understanding of the technology and strategy driving one of Apple’s most significant competitors. He now reports directly to software chief Craig Federighi, with a clear mission to accelerate Apple’s AI development and execution.

The need for a reset became increasingly apparent after the rocky launch of Apple Intelligence in October 2024. The platform’s reception was largely negative, with critics labeling it underwhelming. Early features malfunctioned publicly; a notification summary tool, for example, repeatedly generated false and embarrassing headlines. In one instance, it incorrectly reported that a man accused in a high-profile case had died by suicide. In another, it announced a darts championship winner before the final match had even begun.

These public stumbles were just the surface of deeper internal issues. A delayed and incomplete overhaul of Siri became a significant embarrassment. Reports indicate that just weeks before a planned April launch, senior executives testing the new Siri found that many promised features simply did not work. The launch was postponed indefinitely, leading to class-action lawsuits from customers who had purchased new iPhones expecting a fully functional AI assistant.

Internally, the situation was described as dysfunctional. Communication between AI and marketing teams was poor, budgets were misaligned, and morale suffered. Some employees reportedly referred to Giannandrea’s division mockingly as “AI/MLess.” A wave of AI researchers departed for competitors like OpenAI, Google, and Meta. By early 2025, Giannandrea’s responsibilities had been significantly reduced, with Siri development and other key projects reassigned to other executives.

The scale of the challenge is underscored by Apple’s reported new direction for Siri. The company is now leaning on Google’s Gemini AI models to power the next iteration of its assistant, a remarkable and likely humbling partnership given the historic rivalry between the two firms across numerous product categories.

Apple’s overall AI philosophy has centered on a privacy-first approach, processing most tasks directly on a user’s device using its custom silicon chips. For more complex requests, it uses a system called Private Cloud Compute, which is designed to process data temporarily without permanent storage. While this addresses privacy concerns, it involves trade-offs. On-device AI models are necessarily smaller and less powerful than the massive, cloud-based models used by competitors. Furthermore, Apple’s strict data collection policies mean its models are often trained on licensed or synthetic data, rather than the vast real-world datasets that fuel rivals’ systems.

The central question now is whether this distinctive approach will ultimately prove to be a competitive advantage or a permanent handicap. With Amar Subramanya at the helm, Apple is betting that leadership with insider knowledge of the competition can navigate these complex technical and strategic challenges to finally deliver the sophisticated, reliable AI experiences its users expect.

(Source: TechCrunch)