Elon Musk just scored another win in his race to dominate the artificial intelligence space, and this time it’s not about cars, rockets, or satellites. It’s about government contracts. Musk’s xAI has locked in a deal with the U.S. General Services Administration (GSA) to provide its Grok chatbot to federal agencies, the procurement arm of the federal government announced Thursday. The agreement is effective through March 2027, marking a significant step in Washington’s expanding use of AI for its operations.
Here’s the kicker: price. According to the GSA, agencies can buy Grok models for just 42 cents per organization. That’s less than half of what OpenAI charges—$1 per year—for access to ChatGPT. Government bean counters love a bargain, and xAI is clearly positioning Grok as the budget-friendly choice. The contract covers Grok 4 and Grok 4 Fast, which xAI describes as its “most advanced reasoning models.” Engineers from the Musk-led startup will also provide direct support to federal agencies for implementation, which could give xAI a foothold in the lucrative enterprise IT space.
Agencies will have the option to upgrade to Grok enterprise subscriptions aligned with federal security standards. Those upgrades include expanded features and higher usage limits, which could make Grok an attractive tool for everything from data processing to interagency communication.
But there’s controversy here too. Critics of Grok say the system has a tendency to spit out factually wrong answers, politically biased commentary, and even offensive language. Advocacy groups have flagged several instances where Grok’s responses veered into conspiracy theories or culturally insensitive phrasing. That raises big questions about safeguards—because once you let an AI into federal systems, errors can turn into expensive, embarrassing, or even dangerous mistakes.
The Musk deal is part of GSA’s “OneGov Strategy,” launched in April to streamline procurement and make AI adoption more consistent across federal agencies. Other players already in the program include OpenAI, Google, Anthropic, and even Meta, whose Llama model was just approved earlier this week and made available to agencies at no cost.
So what does this all mean? For Musk, it’s another notch in his belt. For Washington, it’s a gamble—embracing cutting-edge AI while critics are still raising alarms about accuracy, bias, and security. One thing is certain: the AI arms race has officially reached the federal bureaucracy, and Musk just undercut his competition.