Elon Musk is suing OpenAI for betraying its founding mission by pursuing defense contracts. But the same week his lawsuit landed in court, his own company xAI quietly signed an identical agreement with the Pentagon. The contradiction at the heart of Musk's legal crusade just became impossible to ignore.
The Department of Defense announced Friday it had finalized classified AI agreements with seven companies: OpenAI, Google, Microsoft, Amazon, Nvidia, xAI, and the startup Reflection. The deals permit these vendors to deploy their AI systems on classified government networks — a significant expansion of military AI infrastructure. The same announcement declared Anthropic a supply-chain risk, effectively banning the company from Pentagon work it had previously conducted.
Musk's lawsuit, filed earlier this year, accuses OpenAI of abandoning its commitment to developing AI for public benefit in favor of lucrative defense partnerships. The filing specifically cited OpenAI's Pentagon work as evidence of mission corruption. What the lawsuit failed to mention: xAI has been negotiating its own classified deal with the same agency since at least late 2025, according to reporting by The Information.
The timing is damning. Musk's legal team moved for an injunction to block OpenAI from entering defense contracts while simultaneously not disclosing — and presumably not challenging — xAI's parallel pursuit of identical work. Defense attorneys not involved in the case say this creates a fundamental credibility problem. "You cannot simultaneously argue that Pentagon partnerships corrupt AI development and then pursue those exact partnerships through your own company," said one intellectual property lawyer who reviewed the filings. "The court will notice."
OpenAI has defended its defense work as consistent with its stated mission of ensuring AI benefits humanity — noting that military applications can serve democratic institutions. The company did not respond to requests for comment on the lawsuit's changed posture.
The Anthropic exclusion reveals another layer. The company, which powers Claude and has maintained more cautious public statements about military AI, was flagged as a supply-chain risk — language typically reserved for vendors deemed vulnerable to foreign interference. Anthropic has denied any security concerns and called the designation politically motivated. Some analysts suggest the exclusion reflects Anthropic's refusal to agree to liability terms that competitors accepted.
For Musk, the stakes extend beyond the lawsuit. xAI's Pentagon relationship positions the company as a serious defense contractor, competing directly with OpenAI for classified AI work worth billions in government contracts. The lawsuit may serve a dual purpose: damage a competitor's reputation while securing his own company's footprint.
What happens next is straightforward: OpenAI's legal team will almost certainly file a countersuit pointing to xAI's own Pentagon deals. The court will have to decide whether pursuing defense contracts constitutes mission betrayal — and if so, why xAI faces no consequences. Musk's paradox will face judicial scrutiny before the year ends.
The broader picture is clear. American AI companies are racing to embed themselves in defense infrastructure. Every major player — including those founded on anti-military principles — has now signed classified agreements. The distinction Musk once claimed between OpenAI's corruption and his own company's purity has effectively evaporated.