After the loss of about a dozen AI engineers, developing Apple Intelligence features and shipping that improved Siri in 2026 could be more difficult than Apple hoped.
Apple CEO Tim Cook and SVP Craig Federighi recently told employees that AI would be a focus going forward. But a new report claims that the company continues to lose key AI talent to the competition.
“Apple has lost around a dozen of its artificial intelligence staff, including top researchers,” a Financial Times report warns. Meta, OpenAI, xAI, and AI startup Cohere are said to be the companies picking at the bones of Apple’s AI teams.
The report notes that OpenAI has already poached Brandon McKinzie and Dian Ang Yap. Both were foundational models research engineers at Apple.
Machine learning scientist Liutong Zhou was reportedly hired by Chohere. Ruoming Pang, head of Apple’s foundational models team, now works for Meta.
The Facebook owner recently kick-started an AI hiring spree thanks to a $100 million pool for signing bonuses.
Low-hanging fruit
Aaron Sines, director of AI recruiting at Razoroo, told the Financial Times that the loss of Pang is particularly problematic. Pang’s move to Meta has given the industry the belief that Apple’s AI engineers are ripe for the taking, when the offer is right.
Other AI team losses reportedly include Mark Lee, Tom Gunter, Bowen Zhang, and Shuang Ma. All left Apple for Meta.
This comes amid growing competition in the AI space. Apple currently leverages OpenAI’s ChatGPT when Siri can’t handle a question itself, a sign that it still has plenty of catching up to do. Losses of key personnel are unlikely to help.
Apple has already seen its AI tools suffer delays, even before the most recent engineer exodus. A revamped and much-improved Siri experience was announced alongside iOS 18 and is still missing from recent iOS 26 betas.
Federighi, Apple’s software engineering chief, told employees that the delay was because it was trying to bring large language models to existing Siri systems. “We initially wanted to do a hybrid architecture, but we realized that approach wasn’t going to get us to Apple quality,” Federighi said.
As a result, Apple is now back to the drawing board and intends to work on an “end-to-end” rebuild of Siri. But it needs AI engineers to make that happen.
Shallow hiring pool, deep needs
While some employee churn is to be expected, it’s not clear if these departures are at Silicon Valley’s normal high turnover rate. And departures are easy to spot, whereas initial hires are far harder.
As compared to other engineering disciplines that Apple needs, Sines believes that there are only “a thousand, maybe two thousand” engineers in total across the entire hiring body that Silicon Valley is picking from who have top-notch foundational model experience. Apple will hope to hire, and then retain, as many of them as possible.
But with companies like Meta willing to throw huge amounts of money at the best engineers around, that could be difficult. Apple doesn’t have a reputation for being the highest-paying outfit in Silicon Valley — and that might need to change.