OpenAI reportedly disbanded its preparedness team
At the end of last month OpenAI officially disbands the unit that evaluated whether its models posed “serious risks” and designed mitigation strategies. The team, led by Dylan Scandinaro—who left Anthropic earlier this year—shifts its focus to the technical challenge of “recursive self‑improving” AI, while broader safety responsibilities are parceled out to bio‑security, cyber‑security and other functional squads. This restructuring arrives amid a wave of departures that include ethics lead Chloé Bakalar, chief futurist Josh Achiam, and head of safety Johannes Heidecke, signaling a rapid contraction of OpenAI’s internal safety apparatus just as the company prepares for a “massive” public offering.
The move reflects a larger pattern within OpenAI of trimming long‑standing safety programs such as the AGI‑readiness and superalignment groups. Over the past two years the firm has pivoted toward shipping high‑visibility products—ChatGPT, GPT‑4 Turbo, and upcoming multimodal tools—while competitors like Anthropic and Google DeepMind publicly double down on alignment research. Investor pressure to demonstrate revenue growth and market traction appears to outweigh the earlier, more cautious stance that emphasized pre‑emptive risk management. Critics, including former OpenAI researcher Jan Leike, argue that the company is now prioritizing “shiny products” over robust safeguards, a narrative that could shape regulatory and public perception as AI governance debates intensify.
If OpenAI’s safety functions become fragmented across product teams, the company may face gaps in coordinated risk oversight, especially concerning emergent capabilities like self‑improving systems. Regulators could scrutinize the firm’s governance structure during the IPO filing process, demanding clearer accountability mechanisms. Watch for the establishment of any new cross‑functional safety board, external audit commitments, or policy disclosures that attempt to compensate for the dissolved unit. Additionally, Scandinaro’s narrowed remit may produce deeper technical insights but could leave broader societal risks—bias, misinformation, or misuse—less monitored, potentially affecting partner confidence and user trust.
Key Takeaways
OpenAI eliminates its dedicated preparedness team, redistributing risk‑assessment duties to existing product groups.
The restructuring coincides with the exit of key safety and ethics leaders, shrinking the company’s internal alignment capacity.
The shift underscores a strategic emphasis on product rollout and IPO readiness over comprehensive safety governance.
Investors, regulators, and partners will likely monitor how OpenAI fills the oversight gap and whether new safety structures emerge.
About the Source
This analysis is based on reporting by The Verge. Here is a short excerpt for context:
According to the Financial Times, OpenAI disbanded its preparedness team at the end of last month. The job of the preparedness team was to assess if models posed serious risks and develop ways to mitigate those risks. (You know, like the possibility that it could go rogue and hack another company.) According to FT, responsibility has instead been divided up for specific areas like bio and cyber, then moved into existing teams. This is the latest change at the company, which has been in upheaval as it heads towards what is expected to be a massive IPO. Over the last few years, it's slowly torn down its more reach-led model, dissolving its AG … Read the full story at The Verge.Read the original at The Verge
