Musk Lawsuit Cites OpenAI Safety Issues in GPT‑4 India
SkimNews Take
Musk's lawsuit forces OpenAI to defend its safety commitments just as the dissolution of Prism and shelving of Sora thin out the organizational infrastructure those commitments depend on — meaning the courtroom defense and the on-the-ground capacity for safety work are contracting in the same window.
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI: was criticized in Oakland federal court testimony that its product‑focused rollout compromised its AI safety commitment, per former employee Rosie Campbell.
- Rosie Campbell: said she joined OpenAI’s AGI readiness team in 2021, left in 2024 after her team was disbanded, and noted the simultaneous shutdown of the Super Alignment safety team.
- Microsoft: deployed a version of OpenAI’s GPT‑4 model in India via Bing before the model had been evaluated by OpenAI’s Deployment Safety Board, according to Campbell’s testimony.
- Dylan Scandinaro: was hired from Anthropic in February to lead OpenAI’s preparedness team, while the company declined comment on its current AGI alignment approach.
- Sam Altman: was briefly fired by OpenAI’s non‑profit board in 2023 after the India GPT‑4 rollout raised safety concerns, and board member Tasha McCauley testified about his lack of disclosure and conflict‑averse management.
- David Schizer: the former Columbia Law School dean paid by Musk’s team as expert witness, echoed board concerns that OpenAI must prioritize safety over profits and follow formal safety review processes.
Why it matters: The testimony underscores that OpenAI's profit‑driven product push may sideline safety reviews, giving regulators and public advocates a stronger case for stricter oversight while potentially slowing OpenAI's market rollout and limiting its competitive edge and could reshape the balance between AI innovation and safety governance across the industry.
Ask SkimNews
