OpenAI's six-member board holds the power to declare AGI

Microsoft has poured billions into OpenAI, but a tiny group of six people could theoretically cut them off from the ultimate prize at any moment.

According to OpenAI's own governance documents, highlighted over the weekend by developer advocate Logan Kilpatrick on X, the company's six-member nonprofit board holds sole authority to decide when artificial general intelligence has been achieved.

Here is the real kicker for Satya Nadella: once that board declares AGI, the technology is officially excluded from Microsoft's commercial licensing and intellectual property agreements. Those contracts only apply to pre-AGI technology.

The half-dozen deciding humanity's future

OpenAI defines AGI as a highly autonomous system that outperforms humans at most economically valuable work. Since that milestone is notoriously vague, the call comes down entirely to six individuals.

Right now, the board consists of three company insiders and three outsiders. The insiders are chairman Greg Brockman, chief scientist Ilya Sutskever, and CEO Sam Altman. The external directors are Quora CEO Adam D'Angelo, tech entrepreneur Tasha McCauley, and Helen Toner, director of strategy at Georgetown University's Center for Security and Emerging Technology.

All three outside members have had reported ties to the Effective Altruism community, though an OpenAI spokesperson pushed back, telling reporters that none of the board members consider themselves effective altruists and that their interactions are strictly focused on AI safety.

Meanwhile, the business reality creates wild contradictions. Microsoft president Brad Smith recently touted OpenAI's nonprofit setup to argue it is more trustworthy than profit-driven rivals like Meta. Yet The Wall Street Journal reports OpenAI is chasing a $90 billion valuation in secondary share sales, and Altman told the Financial Times that training expenses are huge and he hopes Microsoft keeps funding the massive compute needed along the way.

Why it matters

We think this contract clause exposes the most fragile fault line in tech right now. Microsoft is essentially subsidizing the creation of an intelligence that it legally might not be allowed to commercialize once it actually works.

For enterprise customers and developers betting their stacks on Azure OpenAI services, this introduces weird regulatory and contract risks down the line. If OpenAI decides GPT-5 or a future system counts as AGI, the commercial pipeline could freeze or radically change overnight.

OpenAI's structure looks clever on paper because it shields core safety decisions from shareholder greed. But leaving a multi-trillion-dollar definition in the hands of six people with zero outside appeal process feels legally volatile. If the board ever pulls that trigger, expect one of the messiest corporate showdowns in Silicon Valley history.