MockingbirdNews Logo

Mockingbird News

REAL NEWS NEVER FELT FUNNIER

Categories

OpenAI’s Safety Leads Exit Faster Than Everyone Can Say 'Alignment Crisis'

KEY POINTS

  • Johannes Heidecke departed OpenAI in 2026 as the company merged its Safety and Research teams under Mia Glaese.
  • Several top safety leaders, including Ilya Sutskever and Jan Leike, left in 2024 amid worries OpenAI prioritized products over safety.
  • Competitor Anthropic attracted defectors like Leike and Andrea Vallone, highlighting OpenAI’s ongoing challenges retaining safety experts.

OpenAI’s safety leadership team isn’t just leaking, it’s a full-on evacuation. Johannes Heidecke, who took over Safety Systems in 2021, just waved goodbye amid a company shake-up merging safety with research under Mia Glaese’s watchful eye. He’s the latest casualty in a roster including names like Jan Leike and Ilya Sutskever—both co-founders who quit in May 2024 after declaring that product hype was tripping over safety. Anthropic even stole some defectors like Leike and Andrea Vallone, highlighting OpenAI’s revolving door of safety gurus. Mark Chen called it graceful, but Andrew Curran compared it to the cursed Defense Against the Dark Arts post from Harry Potter, except here the magic is disappearing job titles.

Share the Story

(1 of 3)

Source: Businessinsider | Published: 7/12/2026 | Author: Lakshmi Varanasi

Read the original article →