Grok has Elon Musk's platform, a federal discount, and political proximity. What it still lacks is the institutional trust that turns an AI chatbot into serious enterprise software.
The uncomfortable lesson from Washington is simple: attention is not adoption. xAI put Grok in front of federal agencies at a symbolic price, wrapped it in the language of frontier AI, and still watched buyers lean toward more familiar tools from OpenAI, Microsoft, Google, and Anthropic.
According to a Reuters review published on May 21, federal AI inventory records for 2025 listed more than 400 public government AI use cases that named a specific vendor. Only three involved Grok or xAI. OpenAI-based tools, including ChatGPT, Codex, and Microsoft Copilot, appeared in 234 examples. Alphabet's Gemini and related products appeared in 33. Anthropic's Claude appeared in 26.
That gap matters because Grok was not priced out of the market. The U.S. General Services Administration announced in September 2025 that agencies could access Grok 4 and Grok 4 Fast for 42 cents per organization under its OneGov program, with the offer running for 18 months through March 2027. xAI also promised engineering support for agencies that wanted help putting the models into workflows. The government had a low-cost path to try the product. Most agencies still did not seem eager to build around it.
For entrepreneurs, this is the part worth slowing down on. Consumer distribution can create awareness quickly, but enterprise distribution depends on a different kind of confidence. A federal procurement team is not buying a personality, a viral brand, or a chatbot that performs well in public demos. It is buying security, auditability, contract support, compliance posture, integration depth, and a vendor it can call when something breaks.
That is where Grok's public identity may work against it. Musk has promoted the chatbot as a less constrained, truth-seeking alternative to rivals. That may appeal to some individual users on X, where speed, attitude, and novelty can travel fast. Government departments and regulated companies operate under another set of incentives. They do not want an assistant that surprises them. They want one that behaves predictably inside boring but critical systems.
The Reuters report also noted that Grok's limited federal uses were for relatively low-level work, including first drafts of documents or social media tasks at agencies such as the Office of Personnel Management and the Department of Health and Human Services. That is not meaningless, but it is a long way from becoming embedded infrastructure for fraud detection, case management, classified analysis, or operational decision support.
The SpaceX story now has an AI question
The timing is awkward because xAI is no longer just a side project in Musk's empire. Reuters reported that SpaceX's initial public offering is being pitched partly around an AI growth story, with a proposed valuation around $1.75 trillion and a claimed $26.5 trillion opportunity in AI services for large companies and organizations. Other reports have said the company is targeting a June 12 listing and could raise as much as $75 billion, though the timing and size remain subject to the IPO process.
Those are giant numbers, and giant numbers invite harder questions. If one of the world's largest technology buyers is barely using Grok after being offered near-free access, investors will ask whether xAI's enterprise demand is real or merely assumed. Forbes, reviewing SpaceX's IPO filing, reported that X has around 550 million monthly active users and that 117 million monthly active users engaged with Grok features. That shows reach. It does not prove enterprise conversion.
The filing itself also shows why this is not just a usage story. Forbes reported that SpaceX warned investors about reputational and legal risks tied to Grok features such as spicy mode, including possible misinformation and nonconsensual or exploitative imagery. That kind of disclosure may be standard risk language, but it lands differently when the product is being sold into government and regulated markets. Buyers in those rooms are paid to worry about edge cases before they become public problems.
There is a reason Microsoft Copilot has an advantage in these rooms. It arrives inside systems that agencies and companies already use. OpenAI benefits from that Microsoft channel. Google has Workspace and cloud relationships. Anthropic has made safety and enterprise deployment central to its pitch, even as its government position has become more politically complicated. Grok, by contrast, is still trying to convert a loud consumer presence into institutional permission.
Netskope's enterprise data adds another signal. Its AI Index, current through May 2026, shows ChatGPT and Claude with broad organizational adoption, while Reuters separately cited Netskope analysis showing Grok remained marginal in business use, with far less time spent in the tool than in ChatGPT. The exact measurement will vary by sample, but the direction is clear enough.
This does not mean Grok cannot improve. xAI has talent, compute ambition, and a founder who understands how to force a product into public conversation. The company is also part of broader federal model-testing arrangements, with Reuters reporting earlier in May that xAI, Microsoft, and Google agreed to give the U.S. government early access to new AI models for security checks. That keeps xAI at the table.
But being at the table is not the same as being chosen. The practical takeaway for every AI challenger is that political access and founder celebrity may open doors, but they do not close enterprise sales by themselves. The next phase of AI competition will be won by companies that make buyers feel protected, not merely impressed. Grok has visibility. Now it has to prove it can survive the dull, demanding work of becoming trusted software.
Also read: Brain implants are pushing artificial vision closer to practical use • A $4,000 bounty shows DeFi still prices security too cheaply. • Uber and DoorDash are testing a new delivery merger playbook
