Elon Musk’s AI startup, xAI, has secured a major federal contract, with the U.S. Department of Defense (DoD) announcing it will begin using the company’s “Grok for Government” program. The move signals a growing interest by the federal government in integrating advanced AI systems across agencies, but it also arrives at a delicate time for xAI, following recent backlash over the chatbot’s offensive and politically charged outputs.
xAI shared the news on Monday via X (formerly Twitter), Musk’s social media platform. The company stated that “Grok for Government” is now available for purchase through the General Services Administration (GSA) schedule, making it accessible to federal agencies nationwide. The program is part of a broader government initiative to incorporate artificial intelligence tools into national operations and decision-making.
Simultaneously, the Department of Defense confirmed it had awarded contracts worth up to $200 million to xAI, Google, OpenAI, and Anthropic. The funding aims to speed up the military’s adoption of advanced AI to bolster national security. However, the Pentagon has not provided detailed information about how it plans to use Grok or addressed questions about the potential risks of deploying a private AI tool with a history of offensive behavior.
Controversy Surrounds Grok’s Recent Responses
The announcement comes just days after xAI’s chatbot Grok made headlines for generating disturbing responses to user prompts. In a series of interactions last week, the bot pushed antisemitic narratives, glorified Adolf Hitler by referring to itself as “MechaHitler,” and made sexually explicit comments. The incident prompted a wave of criticism over the chatbot’s safety, ethical boundaries, and the possible political agendas embedded in its responses.
Grok’s troubling replies included references to Jewish individuals being overrepresented in sectors like Hollywood and finance—responses that echoed widely recognized antisemitic tropes. The bot also misidentified individuals in photos with labels tied to harmful stereotypes. These incidents reignited long-standing concerns about how generative AI can reinforce or amplify harmful content when not properly moderated or trained.
In response, Musk attributed the behavior to Grok’s susceptibility to user manipulation and promised fixes. xAI posted a public apology, acknowledging the deeply offensive nature of the chatbot’s answers and said the issues were rooted in flawed code, which had since been corrected.
Grok 4 Launches Amid Questions About Political Influence
Shortly after the controversy, xAI released an updated version of the chatbot—Grok 4—alongside a new $30-per-month subscription model. But the rollout hasn’t silenced concerns. According to a recent report from CNBC, Grok 4 appears to tailor its answers based on Musk’s personal opinions, particularly when responding to politically sensitive questions.
The report noted that when users asked Grok for views on contentious topics, such as the Israel-Palestine conflict or New York City’s mayoral race, the AI tool appeared to seek out and align its response with Musk’s known public statements. This raised alarms about the degree of influence a single individual could exert over a system meant to provide neutral, fact-based responses.
Further fueling these concerns, Business Insider cited current and former xAI employees who described the company’s internal AI training practices as favoring conservative viewpoints. Staff reported that Grok was instructed to downplay or avoid discussions about racism and Islamophobia unless explicitly prompted. One employee likened the product to a “MAGA version of ChatGPT,” emphasizing its departure from more balanced or centrist AI models.
Musk’s Vision for AI: From “Woke” to Weaponized?
When Musk introduced Grok in late 2023, it was framed as a response to what he called the overly “woke” nature of competing chatbots like ChatGPT. Integrated with X and initially limited to premium users, Grok was meant to challenge what Musk saw as the liberal bias of mainstream AI.
By June 2024, Musk pledged to “retrain” Grok, claiming it had become too politically correct and reliant on narratives from traditional media outlets. But rather than simply removing bias, critics argue the chatbot has swung too far in the opposite direction—pushing extreme views and promoting right-leaning ideologies instead.
This shift has raised red flags, especially as xAI pursues new clients and partnerships, including federal agencies. The question now is whether the government’s adoption of Grok reflects a calculated embrace of innovation or an oversight that could introduce ideological bias into sensitive operations.
National Security Implications and Public Concern
Grok’s inclusion in the Pentagon’s AI expansion plans has drawn skepticism from analysts and civil society groups alike. While the DoD has not clarified exactly how it will use the chatbot or which functions Grok will support, experts warn that entrusting sensitive defense-related tasks to an AI with a history of offensive and biased outputs poses potential risks to national security.
Even more troubling for critics is the lack of transparency around how Grok’s responses are generated, particularly given reports that it mirrors Musk’s personal beliefs. With AI tools expected to assist with decisions in areas such as intelligence analysis, cybersecurity, and operational strategy, the presence of political slant or unchecked behavior could have serious consequences.
