Elon Musk’s latest warning about artificial intelligence landed with the force of a movie trailer, but the fight unfolding in an Oakland federal courtroom is much more concrete.
During testimony in his lawsuit against OpenAI, Musk described AI as a “double-edged sword” that could “solve all the diseases and make everyone prosperous, or it could kill us all.”
That sounds like science fiction, but behind the dramatic language sits a very real question for business, technology, and the environment. Who controls the companies building powerful AI, and how much electricity, land, water, hardware, and public trust will be needed to keep that race moving?
A warning in court
Musk is suing OpenAI, CEO Sam Altman, President Greg Brockman, and Microsoft over claims that OpenAI betrayed its original nonprofit mission and shifted toward a profit-driven model.
Reuters reported that Musk is seeking leadership changes and $150 billion in damages from OpenAI and Microsoft, one of the company’s largest investors.
OpenAI rejects that framing. Its lawyers have argued that Musk is trying to regain influence over a company he left in 2018, while also running his own AI competitor, xAI.
The courtroom has already seen some sharp moments. Musk testified that he did not read the “fine print” of a 2017 term sheet connected to OpenAI’s move toward a for-profit structure, saying he had only read the headline.
Why the nonprofit fight matters
OpenAI says it was founded in 2015 as a nonprofit, then created a for-profit subsidiary in 2019 to help scale research and deployment. The company also says that the for-profit business has always been governed and controlled by the nonprofit.
That is the heart of the dispute. Musk says the public-interest mission was pushed aside. OpenAI says the structure was needed because advanced AI takes enormous amounts of money, talent, and computing power.
So what is really at stake? Not just one board seat, one old promise, or one billionaire’s memory of what happened in 2017. It is the model for how a technology that could touch schools, hospitals, businesses, cybersecurity, and critical infrastructure gets built.
The energy question
AI does not float quietly in “the cloud.” It lives in data centers, which need constant power, cooling, chips, backup systems, and grid connections.
The International Energy Agency says global electricity use from data centers is projected to double to around 945 terawatt-hours by 2030. By the agency’s estimates, data center electricity consumption is growing much faster than demand from most other sectors, while accelerated servers driven mainly by AI adoption are projected to grow even faster.
That matters for everyday life. When a region is already dealing with summer heat, air conditioners, and higher electric bills, adding huge new power loads is not a small detail. It can become a local political issue very quickly.
Clean power is not automatic
There is some good news. The IEA says renewables are expected to meet nearly half of the additional electricity demand from data centers over the next few years.
But there is a catch: the same analysis says natural gas and coal together are still expected to supply more than 40% of the additional electricity demand from data centers through 2030. In practical terms, AI growth can still put pressure on emissions unless clean power, storage, and grid upgrades move fast enough.
OpenAI has also made clear that scale is part of its plan. In an April 27 statement about its Microsoft partnership, the company pointed to “scaling gigawatts of new datacenter capacity,” next-generation silicon, and cybersecurity work.
Business risk gets physical
For years, AI risk sounded like something abstract. Bias, misinformation, job disruption, runaway systems. All of that still matters.
But the OpenAI trial shows another layer. The AI economy is tied to physical assets, including power plants, chips, transformers, cooling systems, and data center campuses. The IEA saysAI data centers are already facing bottlenecks involving energy supply chains, grid connections, advanced chips, and regulatory approvals.
That turns AI governance into an environmental question, too. A company’s legal structure can affect who makes investment decisions, who answers to the public, and who carries the burden when infrastructure strains local communities.
What comes next
The trial began on April 28 in Oakland and is expected to last several weeks, with testimony expected from Altman, Brockman, and Microsoft CEO Satya Nadella.
The jury will not decide the entire future of AI, nor will it settle every fear about safety, competition, or energy demand. But it may influence how one of the world’s most important AI companies is governed at a moment when the industry is growing from software into something much larger.
The trouble is, the clock is moving fast. Data centers can be built in a few years, while power grids and permitting systems often move much more slowly. That gap is where the environmental stakes begin to show.
The official announcement about audio access to the trial was published on the U.S. District Court for the Northern District of California.
