Do we like xAI?
Elon Musk's AI company, maker of the Grok chatbot. Now part of SpaceX and branded SpaceXAI. Based in Palo Alto, California.
Our score: 2 out of 10. We wouldn't back them. Grok's deepfake scandal and the unpermitted Memphis gas turbines are serious harms, and little good offsets them.
xAI was founded by Elon Musk in 2023 and shipped its first Grok model that November. The pitch was a chatbot with a rebellious streak that would answer questions other AIs wouldn't. It bought X in 2025, so the chatbot and the social network ended up under one roof.
In Feb 2026 SpaceX bought xAI in an all-stock deal that Reuters reported valued xAI at $250bn. xAI is now a wholly owned SpaceX subsidiary and has been rebranded SpaceXAI. SpaceX listed on Nasdaq, so the only way to own a piece of xAI is through SpaceX shares. Musk keeps control.
The big scandal is Grok's image tools. Early in 2026 they were used to make non-consensual sexualised images of real people, and researchers said some showed children. Ofcom in the UK and the European Commission opened formal investigations into X over it. Ireland's data regulator opened a privacy inquiry.
Then there's Memphis. xAI ran dozens of gas turbines at its supercomputer site without air permits. The Southern Environmental Law Center and the NAACP threatened to sue under the Clean Air Act. University of Tennessee researchers found the turbines added to local air pollution. Regulators later granted permits for 15 turbines.
The product itself keeps moving fast, with Grok 4.7 out in September 2026. Musk has admitted that running Grok Bot costs far more than it earns. That is a business problem rather than a moral one. Because of the deepfake and pollution records, we score it low.