The US Department of War (formerly Department of Defence) has announced an agreement with Elon Musk’s startup xAI to integrate its Grok chatbot into military systems.
The partnership follows a $200 million deal between xAI and the Pentagon earlier this year to develop an “AI aresnal” to “address critical national security challenges”.
It raised concerns at the time relating to controversies surrounding Grok, as well as the amount of access Mr Musk and his company might get to sensitive government data.
US Senator Elizabeth Warren wrote a letter in September to Secretary of War Pete Hegseth warning that Grok was “generating erroneous outputs and misinformation” and “offensive and antisemitic posts”.
The latest agreement, announced on Monday, will allow all military and civilian personal at the department to use the artificial intelligence to handle controlled unclassified information – known as Impact Level 5 (IL5) – in early 2026.
“This move builds on the rapid deployment of cutting-edge AI across the Department’s 3 million military and civilian personnel,” a statement from the US Department of War read.
“The War Department will continue scaling an AI ecosystem built for speed, security, and decision superiority. Newly IL5-certified capabilities will empower every aspect of the Department’s workforce, turning AI into a daily operational asset.”
The deal will also see users within the Department of War gain access to another of Mr Musk’s platforms, X (formerly Twitter), for “real-time global insights” that give a “decisive information advantage”.
Grok has faced a multitude of controversies since launching in November 2023, with Mr Musk originally pitching it as an edgy alternative to “politically correct” AI chatbots like OpenAI’s ChatGPT.
Earlier this year, an update saw Grok praise Adolf Hitler, share antisemitic tropes and endorse a second Holocaust. It also referred to itself as “MechaHitler”, referencing a robotic version of the Nazi dictator from the video game Wolfenstein 3D.
Offensive comments made about political figures saw a Turkish court block access to the the AI bot in July on the grounds of national security and public order.
Grok has also been criticised for a perceived ideological bias, which often appears to reflect the viewpoints of Mr Musk.
In a statement following the Department of War’s announcement, xAI said the partnership would demonstrate Grok’s capabilities within “critical mission” use cases.
“Through an ongoing, long-term partnership with the DoW and other mission partners, xAI will make available a family of government-optimized foundation models to support classified operational workloads,” the company said.
“These new partnerships are part of our longstanding support of the United States government and xAI’s mission to bring the best tools and technologies available in industry to benefit our nation.”
