Apple's AI Search Chief Ke Yang Departs to Join Meta

Apple has lost another key artificial intelligence leader to Meta, as Ke Yang—who was recently promoted to lead the company's AI search efforts—has departed just weeks after taking on the role.

Yang headed Apple's Answers, Knowledge, and Information (AKI) team, which is responsible for developing a ChatGPT-style web search function for Siri. His exit follows that of his predecessor, Robby Walker, who also left the leadership of the AKI group only a few weeks prior. With Yang's departure, the AKI team now reports to Benoit Dupin, a deputy reporting to John Giannandrea, Apple's senior vice president of machine learning and AI strategy.

Mounting Pressure on Apple's Ambitious Timeline

This timing increases the pressure on Apple's planned major Siri overhaul, scheduled for March 2026. This redesign, based on a large language model, aims to make Siri more conversational and able to pull live information from the web. Internally called "LLM Siri," the new system will include a "World Knowledge Answers" engine that creates AI-powered, multimodal summaries blending text, images, and on-device data—similar to features offered by ChatGPT and Google's AI Overviews.

Yang's move to Meta is part of a wider pattern of talent attrition within Apple's AI division. Over the past year, Meta has hired around a dozen members from Apple's AI and machine learning teams. This group includes Ruoming Pang, who established Apple's Foundation Models team; Jian Zhang, who worked on robotics AI; and Frank Chu, who focused on AI cloud infrastructure and search.

These departures coincide with Meta's expansion of its Superintelligence Labs and acceleration of its generative AI projects. Industry analysts partly attribute this talent shift to Meta's competitive compensation packages and aggressive stance in the AI field. Notably, there hasn't been a similar wave of senior AI executive departures reported at Google, OpenAI, or other leading competitors during this same period.

Apple formed the AKI team earlier in 2025 to help Siri compete with new AI search tools. The team's goal is to enable Siri to fetch real-time information online and answer complex questions in one go, evolving the assistant from simple voice commands into a true "answer engine."

The technical framework for the new Siri relies on three core LLM-driven components: a planner to understand user intent and context, an AI search engine to gather content from the web and user devices, and a summarizer to produce coherent answers. Apple is testing its own AI models alongside external ones, including a tailored version of Google's Gemini AI for summarization, while Apple's Foundation Models handle on-device data to maintain user privacy.

Growing Concerns Over Leadership Stability

The ongoing leadership changes raise doubts about Apple's ability to hit its March 2026 launch date for the new Siri. The company has already delayed several highly anticipated Siri features due to technical constraints and is now shifting entirely to a system built on large language models. This upcoming release is poised to be Apple's most significant digital assistant upgrade, aiming to turn Siri into a proactive, conversational AI capable of natural, extended dialogues and managing multi-step tasks.

Beyond Siri, Apple intends to roll out AI search and answer capabilities to other platforms like Safari and Spotlight, progressing toward a unified AI search experience across its devices. The company continues to internally emphasize that the Siri revamp and AI innovation are top priorities. However, the recent leadership exits and organizational shifts have increased skepticism about its capacity to catch up with the frontrunners in AI-powered assistants.

Yang's departure is the latest challenge for a project that is central to Apple's strategy to rival generative AI assistants from OpenAI, Google, and Perplexity. With the AKI team under new interim leadership and facing a firm March 2026 deadline, Apple needs to solidify its AI organization while preventing further key talent from moving to competitors like Meta.