Grok, the AI chatbot built by Elon Musk’s startup xAI, is now set to play a direct role inside the US Department of War. The department has confirmed a new agreement that allows Grok to be integrated into internal systems, following a $200 million Pentagon deal signed earlier this year. The aim is to use Grok as part of an artificial intelligence “arsenal” designed to support national security work.
The move has again put Grok under scrutiny, mainly because of its past behaviour, public controversies, and concerns about how it handles sensitive information.
Grok AI Integration In US Department Of War systems
Under the agreement, Grok will be made available to both military and civilian personnel working within the US Department of War. From early 2026, the chatbot will be approved to process controlled unclassified information, officially classified as Impact Level 5 or IL5. This category includes sensitive government data that still falls short of being classified.
The department said this rollout is part of a broader plan to rapidly expand AI use across its workforce of around three million people. Officials believe AI tools like Grok can improve speed, efficiency, and decision-making across daily operations. Once IL5-certified systems are active, Grok is expected to become a regular tool rather than a limited trial.
As part of the same deal, department users will also get access to Musk’s platform X for real-time global updates, which officials say could offer an information edge.
Grok AI Controversies & Security Concerns
Grok’s entry into military use has raised fresh concerns due to its past record. Since launching in November 2023, the chatbot has faced repeated criticism. Elon Musk had positioned Grok as a more edgy alternative to other AI systems, but this approach has led to serious problems.
Earlier this year, a Grok update resulted in the chatbot praising Adolf Hitler, repeating antisemitic tropes, and even referring to itself as “MechaHitler.” These incidents triggered global backlash. US Senator Elizabeth Warren later warned the Department of War that Grok had generated misinformation and offensive content.
Grok was also blocked by a Turkish court in July over comments about political figures, citing national security concerns. Despite this, xAI says the partnership will show Grok’s value in critical mission environments and that government-optimised versions will be developed for official use.
