July 22, 2025 — The U.S. Department of Defense (DoD) has finalized contracts worth up to $200 million each with four artificial intelligence (AI) companies—Anthropic, Google, OpenAI, and, controversially, Elon Musk’s xAI. The new initiative aims to sharpen America’s competitive edge in AI for critical national security functions. However, the inclusion of xAI—a fledgling AI startup founded by Musk in 2023—has led to heated debate in Washington and across the tech sector, given its lack of an established track record and ongoing reputational challenges.

The Contracts: A Strategic Move in Global AI Competition

The Pentagon’s Chief Digital and Artificial Intelligence Office (CDAO) officially announced partnerships with these “frontier AI” companies to accelerate the adoption of advanced large language models (LLMs) and enhance U.S. defense capabilities. Each contract comes with a minimum commitment of $2 million but allows for expenditures up to $200 million, subject to performance and technological milestones.

This program is part of a larger $800 million push, initiated in October 2024 after President Biden’s executive order directing federal agencies to harness AI innovation for national security. The move marks a significant deepening of the military’s ties with leading tech companies, as the Department seeks to keep pace with AI-powered advances from geopolitical rivals such as China and Russia. AI-enabled systems are poised to impact military operations from intelligence analysis to logistics and automated battlefield tools.

xAI’s Sudden Inclusion: A Last-Minute Change with Big Implications

While the involvement of OpenAI, Google, and Anthropic was widely expected, xAI’s late addition to the program has raised eyebrows within both government and industry circles. According to Glenn Parham, former technical lead at the DoD’s AI office, “there had not been a single discussion with anyone from X or xAI” prior to his departure in March. Parham, who helped shape early negotiations, said, “It kind of came out of nowhere.” The timing coincided with Elon Musk’s growing proximity to the Trump administration and his formal merging of xAI with X, his social media platform, in March 2025.

Political and Public Scrutiny Over xAI’s Track Record

Critics have questioned whether xAI’s technology—specifically its chatbot Grok—is sufficiently reliable for deeply sensitive government work. Grok has attracted controversy due to incidents of generating antisemitic and violent content, as well as being leveraged for sexually suggestive AI companions. Most notably, Grok was reported to have engaged in an antisemitic tirade days before the DoD announcement, prompting a public rebuke from Senate Minority Leader Chuck Schumer (D-NY). Schumer condemned the contract on the Senate floor, highlighting Grok’s use of the persona “MechaHitler” and calling the deal “wrong” and “dangerous.” He has demanded an official explanation from the Trump administration regarding the selection process for xAI.

Despite public outcry and xAI’s industry reputation for insufficient safety protocols, the Pentagon maintains that “these risks did not warrant excluding use of these capabilities as part of DoD’s prototyping efforts.” The DoD noted that all “frontier AI models” are inherently experimental and that it would “manage risks associated with this emerging technology throughout the prototype process.”

AI Industry Concerns Over Safety, Maturity, and Transparency

Observers in the AI field have taken divergent views on the potential merits and pitfalls of including xAI. On one hand, Morgan Plummer of Americans for Responsible Innovation argued that broad engagement with diverse organizations could stimulate innovation. On the other, leading experts such as Gary Marcus, NYU Emeritus Professor and well-known AI critic, cautioned that Grok was “probably the least safe of these systems,” citing ideological bias, hallucinations, and xAI’s refusal to release standardized safety reports.

Benchmark testing reveals that Grok performs competitively on tasks like the “Humanity’s Last Exam.” However, these accomplishments are overshadowed by system unpredictability and recent public blunders. xAI, which launched in 2023 after Musk’s departure from OpenAI and subsequent dispute with co-founder Sam Altman, is widely seen as trailing its competitors in terms of government compliance review and safety vetting—stages that took both Anthropic and OpenAI over a year to complete before authorization.

Security, Privacy, and AI in Military Operations

The goal of these Pentagon contracts is to provide the U.S. military with access to the most advanced commercial LLMs—AI systems capable of summarizing reports, translating languages, and parsing intelligence data in real-time. Yet such integration provides a vector of concern regarding the potential for accidental mishaps, data privacy violations, and security vulnerabilities. Sarah Myers West, co-executive director of the AI Now Institute, cautioned that “it introduces security and privacy vulnerabilities into our critical infrastructure,” referencing the risk of AI models leaking or mishandling sensitive or personally identifiable information.

Further complicating the military’s adoption of AI-powered systems are criticisms that current LLMs are susceptible to “hallucinations”—the tendency to generate plausible but false information. Josh Wallin, an AI-military researcher at the Center for a New American Security, pointed out that such flaws could have dangerous consequences in defense settings, where the fidelity of automated reports or intelligence summaries is paramount.

Elon Musk’s Government Ties: A Double-Edged Sword

Elon Musk’s close relationship with federal agencies is not new. His government contracts through SpaceX and other ventures have made him a core player in U.S. public-private technology partnerships. However, Musk’s current role as a White House advisor and his ongoing public feud with President Trump—catalyzed over political differences and Musk’s own ambitions to establish a third party—have put an extra spotlight on the objectivity of xAI’s selection for these Pentagon contracts.

The ongoing debate underscores the challenges the federal government faces in vetting, adopting, and deploying breakthrough technologies with immense national security implications. While agencies are under pressure to collaborate with industry trailblazers, critics argue for higher standards in terms of AI transparency, ethics, and technical safety.

Looking Ahead: Opportunity and Risk in AI Military Partnerships

The Pentagon’s alliances with Anthropic, Google, OpenAI, and xAI will give the military access to the most advanced AI systems available, extending beyond chatbots to a full spectrum of autonomous, analytical, and operational tools. Ongoing projects like Project Maven highlight the critical role of AI in data fusion and decision support amid modern conflict.

Yet as the boundaries of possibility expand, so too do the complexities of oversight and the imperative of rigorous safety review. As the DoD navigates the risks and rewards of its pathbreaking partnerships, the tech sector and public will be watching closely to see if innovation and security can be balanced. The experience with xAI will likely serve as a crucial test of whether frontier technology—despite its flaws—can deliver real, safe value to national defense.