Grok AI Struggles to Gain Traction in Government Agencies
New data from a Reuters report reveals that Elon Musk's AI chatbot Grok is barely used by the U.S. federal government. Despite Musk's claims of building a world-class "truth-seeking" model, Grok appears in only three out of over 400 government AI use cases where vendors were named. In contrast, OpenAI appeared in more than 230 examples, while Google and Anthropic each showed up dozens of times.
Government AI Adoption: Grok vs. Competitors
When Grok did appear in government records, it was used only for basic tasks such as document drafting and social media management. The Election Assistance Commission used it twice for routine administrative work, and the Department of Energy ran a pilot at Lawrence Livermore National Laboratory for document summaries and general research.
Meanwhile, Reuters found 140 entries involving Microsoft and OpenAI, at least 10 for Anthropic, and dozens for Google's Gemini. Even in smaller, more ambitious government AI projects, Grok barely registered.
Why Government Workers Prefer Other AI Models
An unnamed Pentagon source told Reuters the reason is simple: Grok "is just not the best model out there." Staffers tend to prefer Google's Gemini or Anthropic's Claude. Public AI leaderboards support this, with Anthropic, Google, and OpenAI dominating top ranks while Grok rarely cracks the top 10 outside niche categories.
SpaceX IPO Relies on Grok's Success
This weak government adoption is especially awkward for Musk because SpaceX recently absorbed xAI. The rocket venture's IPO filing puts Grok at the center of its investor pitch, claiming a $28.5 trillion total addressable market — almost entirely from AI and enterprise AI, not rockets.
Musk's Strong-Arm Tactics for Grok Sales
According to Reuters, Musk has reportedly pressured banks into buying Grok subscriptions if they want to participate in SpaceX's IPO. However, if Grok underperforms in enterprise settings, these deals may offer only a short-term fix.
Grok's Performance Issues and Controversial Outputs
Musk recently admitted xAI used OpenAI's models to train Grok via a process called distillation — standard when using your own models, but controversial with a rival's system. Meanwhile, Grok's public version has generated offensive and conspiratorial content, including praising Adolf Hitler and creating nonconsensual deepfakes. SpaceX's own IPO filing warns that Grok's "spicy" modes carry "heightened risks" including reputational damage and lawsuits.
Can Grok Compete in the Enterprise AI Market?
Grok's poor government adoption suggests it may struggle to win over enterprise customers. While Musk has positioned Grok as a less censored alternative to ChatGPT, its loose evidentiary standards and track record of offensive outputs make it a risky choice for businesses.
Key Takeaways for AI Investors
- Government data shows minimal Grok adoption — just 3 of 400+ cases
- Competitors dominate — OpenAI, Google, and Anthropic appear in hundreds of use cases
- SpaceX IPO hinges on AI valuation — but Grok's performance and reputation are major risks
As Musk pushes Grok into the enterprise, the question remains: can a chatbot that struggles in government agencies and generates controversial outputs find a home in the corporate world?
Comment