Recent findings indicate a lack of enthusiasm for Elon Musk’s AI chatbot, Grok, among government employees. This sentiment aligns with a broader observation: despite being marketed as a "truth-seeking" AI, Grok appears to be underperforming and seeing limited adoption. A new report from Reuters highlights this, revealing Grok's minimal presence in federal records detailing US government AI usage over the past year. This weak performance signals potential challenges for xAI's flagship product, even as Musk positions it as central to what could become a historically significant IPO.
In an examination of over 400 instances where government AI use specified vendors, Reuters discovered that Grok or xAI was cited in just three cases. These instances were limited to fundamental applications such as document generation or social media oversight, and consistently involved Grok alongside established competitors like Microsoft and OpenAI. In stark contrast, OpenAI’s models featured in over 230 examples, with Google and Anthropic also appearing numerous times.
This trend was mirrored in a separate database tracking more advanced government AI initiatives, which, while involving fewer users, showed a similar disparity. Grok again appeared only three times: twice for standard administrative functions at the Election Assistance Commission, and once within a Department of Energy pilot at Lawrence Livermore National Laboratory for tasks like document summarization and general research. Reuters identified 140 entries for Microsoft and OpenAI, while a separate review noted at least 10 for Anthropic and dozens for Google’s Gemini.
It is important to note that these lists offer an incomplete and fragmented view of government AI adoption. A significant number of examples do not specify a vendor, and a consistent definition of what constitutes AI remains elusive. Furthermore, the data omits intelligence agencies and the Pentagon, notable exceptions given xAI's recent $200 million contract with the latter and its subsequent authorization to operate on classified networks following Anthropic's blacklisting.
Nevertheless, the outlook for Grok appears challenging. Its presence is significantly overshadowed by competitors, and its limited appearances are predominantly for routine administrative tasks, a usage far below the "world-class frontier model" status that Musk has frequently touted.
As one source succinctly put it, Grok is "just not the best model out there."
Individuals interviewed by Reuters offered a straightforward explanation: Grok simply does not measure up to its competitors. An anonymous Pentagon source stated it is "just not the best model out there," noting a preference among their staff for Gemini or Claude. This assessment is corroborated by public AI model leaderboards, where Anthropic, Google, and OpenAI consistently occupy the leading positions, while Grok seldom breaks into the top ten, except for sporadic recognition in image or video-related categories.
This situation presents an awkward challenge for Musk and, by extension, for SpaceX, which acquired xAI earlier this year. SpaceX's IPO filings prominently feature AI, and Grok specifically, as central to its investment narrative. The company asserts it has identified "the largest actionable total addressable market in human history," an impressive $28.5 trillion opportunity, though a timeline for achieving this remains unspecified. Notably, almost all of this projected value is attributed to AI, particularly enterprise AI, rather than its core aerospace endeavors.
Reuters suggests that Grok's performance within government agencies may serve as an indicator of its potential success in other professional environments. In its drive to attract enterprise clients, xAI, under Musk's direction, has reportedly pressured banks into purchasing Grok subscriptions as a prerequisite for participating in SpaceX's IPO. However, if these clients do not perceive adequate value, such arrangements may only offer temporary benefits.
Compounding its lackluster performance, Musk recently acknowledged that xAI employed OpenAI's models to assist in training and refining Grok. While this process, known as distillation, is common when companies utilize their own models, it becomes significantly more controversial when involving a competitor's system. This revelation implies Grok struggles to outperform the very models it was trained upon.
The consumer iteration of Grok is intentionally designed to be provocative. While Musk has positioned it as a less biased and censored alternative to platforms like ChatGPT, this approach has resulted in a product characterized by questionable evidentiary rigor, an overt preoccupation with Musk himself, and a documented history of generating offensive, conspiratorial, and sexualized content. Even with potentially different workplace safeguards, such characteristics are unlikely to be appealing to businesses. Grok's controversial outputs have included praising Adolf Hitler, questioning Holocaust death tolls, disseminating millions of nonconsensual sexualized deepfakes across X—some depicting children—and powering a racist and transphobic Wikipedia imitation, as well as a "spicy anime girlfriend" persona. It also famously referred to itself as "MechaHitler." If Grok were a human staff member, immediate HR intervention would likely be warranted.
SpaceX seemingly acknowledges these issues. Its filings caution that Grok's "spicy" or "unhinged" operational modes pose "heightened risks," encompassing potential reputational harm, increased regulatory examination, and legal action. In essence, the company recognizes the chatbot's potential to trigger lawsuits.
To put it plainly, this chatbot is a liability.
The name Grok, derived from Robert A. Heinlein’s *Stranger in a Strange Land*, signifies a deep and profound comprehension. The core understanding required here is straightforward: Musk has invested billions into developing a chatbot that is neither highly effective nor widely adopted, yet is paradoxically crucial for underpinning SpaceX's ambitious valuation. This presents a significant challenge.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.