Apple AI Search Chief Ke Yang Leaves for Meta

Apple has lost another senior artificial intelligence executive to Meta, with Ke Yang—the recently appointed head of the company’s AI search initiative—departing just weeks after his promotion.

Yang led Apple’s Answers, Knowledge, and Information (AKI) team, the group tasked with building a ChatGPT-like web search capability for Siri. His exit follows the departure of his predecessor, Robby Walker, who also left the AKI leadership role only weeks earlier. With Yang’s resignation, the AKI team now reports to Benoit Dupin, a deputy under John Giannandrea, Apple’s senior vice president of machine learning and AI strategy.

Pressure Building on Apple’s Big Plans

The timing compounds pressure on Apple’s planned March 2026 Siri overhaul, a large language model–based revamp designed to make Siri more conversational and capable of pulling real-time information from the web. The new system, internally called “LLM Siri,” will feature a “World Knowledge Answers” engine that generates AI-powered, multimodal summaries combining text, images, and local data—similar to capabilities offered by ChatGPT and Google’s AI Overviews.

Yang’s move to Meta is part of a broader talent drain affecting Apple’s AI division. Meta has hired approximately a dozen members of Apple’s AI and machine learning team over the past year, including Ruoming Pang, who founded Apple’s Foundation Models group; Jian Zhang, who worked on robotics AI; and Frank Chu, focused on AI cloud infrastructure and search.

The departures come as Meta builds up its Superintelligence Labs and accelerates generative AI development. Industry observers attribute the talent migration partly to Meta’s compensation packages and aggressive positioning in the AI race, while no similar wave of senior AI executive exits has been reported at Google, OpenAI, or other major competitors during the same period.

Apple’s AKI team was established earlier in 2025 to help Siri compete with emerging AI search products. The group’s mission is to enable Siri to retrieve live information from the web and answer complex queries in a single interaction, moving the assistant from basic voice commands to functioning as an “answer engine.”

The technical architecture for the new Siri includes three core LLM-powered components: a planner to interpret user intent and context, an AI search engine to retrieve content from both the web and user devices, and a summarizer to generate coherent responses. Apple is testing its own AI models alongside external options, including a customized version of Google’s Gemini AI for summarization, while Apple’s Foundation Models handle on-device data to preserve privacy.

More Questions About Apple’s Leadership Instability

The leadership instability raises questions about whether Apple can meet its March 2026 launch target for the revamped Siri. The company has already postponed much-anticipated Siri features due to architectural limitations and is now transitioning entirely to a large language model–based system. The upcoming release would represent Apple’s most ambitious digital assistant upgrade, transforming Siri into a proactive, conversational AI capable of natural, multi-turn conversations and multi-step task handling.

Beyond Siri, Apple plans to extend AI search and answer features to other platforms including Safari and Spotlight, moving toward a device-wide AI search experience. The company continues to assure internally that the Siri revamp and AI innovation remain top priorities, but recent exits and organizational changes have fueled doubts about its ability to close the gap with leaders in AI-powered assistants.

Yang’s departure marks the latest setback for a project central to Apple’s strategy to compete with generative AI assistants from OpenAI, Google, and Perplexity. With the AKI team now under new interim leadership and facing a March 2026 deadline, Apple must stabilize its AI organization while defending against further talent losses to rivals like Meta.