Carnegie Mellon AI expert Russ Salakhutdinov hired as a new director of AI research at Apple, says he'll continue to do work at CMU
Russ Salakhutdinov tweets he's now a director of AI research — Apple has struggled to attract top AI talent due to secrecy
Context & Ripple Effects
In 2016, Apple's answer to its secrecy-driven recruiting problem was Russ Salakhutdinov, a Carnegie Mellon researcher brought in as director of AI research while keeping his university post — a hybrid arrangement that let Apple buy credibility it couldn't attract outright. The company had already turned to outside credibility once before, hiring John Giannandrea away from Google in 2018 to run machine learning and AI strategy.
The decade since reads as a loop: each marquee hire has been followed by attrition. Ali Farhadi left for the Allen Institute in 2023, Samy Bengio was hired from Google in 2021 after colleagues' firings, and by early 2026 Apple was losing Siri leadership and multiple researchers to Meta and DeepMind. The Salakhutdinov hire is the opening data point in that cycle.
First-order effects
- Apple gains a named AI research director at a moment when its secrecy culture is explicitly cited as a barrier to attracting top talent — the part-time CMU arrangement is the concession that closes the deal.
- Carnegie Mellon keeps Salakhutdinov's affiliation, preserving its research profile while effectively lending talent to a corporate lab.
Second-order effects
- Google, already Apple's rival and its proven talent pool via Giannandrea and later Bengio, faces a competitor willing to offer dual academic-corporate roles to win recruits its own culture repels.
- Academic labs like CMU become de facto feeder systems whose graduates and faculty can be recruited without full commitment — raising the price universities can extract for cooperation.
Third-order effects
- If the pattern holds, Apple's AI organization becomes structurally dependent on a revolving door: high-profile external hires offsetting continuous outflow toward Meta and DeepMind, rather than building durable internal bench strength.
- Secrecy itself becomes a measurable recruiting tax — companies with open research cultures hold a persistent advantage in the market for senior AI researchers, forcing closed firms into costlier or more flexible arrangements.
The trend: Apple's AI talent strategy has spent a decade oscillating between headline external hires and steady attrition to Meta and DeepMind, because its secrecy culture structurally conflicts with the open-research norms top AI researchers expect.