Sources reveal that sidelining of Mustafa Suleyman, co-founder of DeepMind, from management responsibilities in 2019 was because he bullied staff
Google's parent, Alphabet, has waded through one controversy after another involving its top researchers and executives in the field
Context & Ripple Effects
The WSJ report closes a two-year-old mystery: when DeepMind co-founder Mustafa Suleyman was placed on leave in August 2019, the reason was never stated — sources now attribute his removal from management responsibilities to bullying of staff. It lands on an Alphabet already worn down by repeated controversies involving its top researchers and executives.
The arc since then has been one of steady estrangement: DeepMind's push for autonomy from Google was rebuffed, Suleyman exited to Greylock Partners in early 2022, and DeepMind separately overhauled policies after accusations it mishandled sexual misconduct allegations. A conduct finding against a co-founder reframes all of it as a governance problem, not just a personnel one.
First-order effects
- Alphabet and DeepMind face renewed questions about why a co-founder's conduct took roughly 18 months to surface publicly, and whether internal handling of complaints met the standard the lab later promised in its policy changes.
- Suleyman's standing as a public-facing AI ethics figure — built on the activist profile from his DeepMind founding years — is directly undercut by the reported reason for his sidelining.
Second-order effects
- DeepMind's culture and complaint-handling become a recruiting liability precisely as it fights to retain independence inside Alphabet; rivals can pitch stability and freedom against a lab managing founder-conduct fallout.
- The episode hands ammunition to critics of Google's acquisition-era promises, feeding the broader scrutiny of how the lab governs itself under corporate ownership.
Third-order effects
- If the pattern holds, founder conduct becomes a durable variable in AI-lab valuations and talent flows — visible later when Suleyman moved to pitching Google DeepMind recruits on Microsoft AI roles with higher pay and more freedom, turning a former insider into a direct competitor for the same people.
The trend: Frontier AI labs are being judged less on research output alone and more on governance and founder conduct, which increasingly shape their ability to keep talent and autonomy.