Judge signals she may block OpenAI's nonprofit-to-for-profit conversion in Musk lawsuit

Summary

A federal judge signaled she would likely block OpenAI's conversion from nonprofit to for-profit structure if Elon Musk can establish legal standing in his lawsuit. In a ruling this morning, the judge wrote that if Musk's donation to OpenAI gives him standing to sue, she would find it "very very likely" she'd want to enjoin the entire $100 billion conversion. She cited foundational commitments made by Sam Altman and Greg Brockman to avoid using OpenAI as a vehicle for personal enrichment. The judge stated there would be "no inequity" in preserving OpenAI's nonprofit corporate form if the process moves expeditiously.

The core legal question remains whether Musk has standing at all, a high bar. OpenAI's original Delaware and California incorporation documents explicitly state the company was "organized exclusively for charitable and educational purposes and not organized for the private gain of any person," which forms the backbone of Musk's legal argument. Observers assess the case as roughly even odds or slightly favoring Altman and OpenAI. The legal timeline will likely extend years, potentially outlasting near-term AI breakthroughs.

SoftBank has committed $3 billion to spending on OpenAI agents this year alone. That figure strains credulity given the company's apparent agent spending to date is under $100 million. The required growth trajectory would demand roughly $250 million per month in agent costs, or $8 million daily. The commitment appears intertwined with SoftBank's own fundraising strategy. Masayoshi Son is using OpenAI's revenue projections to lever debt for additional investment in the company, creating a circular dynamic where Altman uses SoftBank's spending commitments to raise capital elsewhere.

OpenAI sits at $4 billion ARR and is projected to reach approximately $11 billion by year-end. Altman has pitched agents to investors at price points between $2,000 and $20,000 per month for different tiers: high-income knowledge workers at the low end, software development agents at up to $10,000, and PhD-level research agents at up to $20,000. OpenAI expects 20 to 25 percent of company revenue to eventually come from agents.

The pricing presents a credibility challenge. Devon AI, built atop OpenAI's models, costs $500 per month and is reportedly effective in organizations. OpenAI's agents would need to be 4 to 40 times more capable than Devon to justify the price premium, or risk customers choosing the cheaper alternative. The agent pricing likely serves as a key narrative tool for fundraising, allowing Altman to project substantial new revenue streams beyond core consumer subscriptions to justify OpenAI's aggressive forward valuation multiples.

Ilya Sutskever's startup Safe Superintelligence is in talks to raise funding at a $30 billion valuation, up from a $5 billion valuation previously. SSI raised roughly $2 billion including commitments from Sequoia and Andreessen Horowitz. The company operates under strict secrecy. Candidates leave phones in Faraday cages before interviews, and the website contains only a 223-word mission statement. Sutskever is hiring promising technologists he can mentor rather than experienced Silicon Valley names likely to job-hop. SSI has committed to releasing no products until it develops superintelligence.

Sequoia and Andreessen Horowitz are also invested in Elon Musk's xAI, creating potential conflicts that are brushed aside as routine in the foundation model race.