Rosie Campbell Testimony: OpenAI Disbanded AGI Safety Team as Profits Took Priority
The Testimony
Rosie Campbell, a former OpenAI board member, testified in the Musk vs. OpenAI trial that the company shifted from being mission-driven to "more like a product-focused organization." She was part of the AGI readiness team starting in 2021 and left in 2024 when the team was disbanded.
What She Said
Campbell testified that building super-smart models without proper safeguards "would not fit with the mission she originally signed on for." Her testimony suggests that commercial pressures may be winning out over safety commitments at OpenAI.
The Context
Campbell was part of OpenAI AGI readiness team and the Super Alignment group, both of which were disbanded in 2024. Musk case argues that this is part of a pattern of sidelining safety experts in favor of product development and commercialization.
Why This Matters
- Raises questions about whether AI labs can keep safety promises under competitive pressure
- Suggests that commercial interests may override safety commitments
- Highlights the tension between rapid development and responsible AI
The Broader Picture
This testimony comes at a time when AI capabilities are advancing rapidly. If the most prominent AI labs cannot maintain their safety commitments under competitive pressure, the industry needs stronger mechanisms to ensure responsible development. This could mean external oversight, industry standards, or regulatory requirements.