The US government has acknowledged that Elon Musk’s AI tool Grok was used in military operations targeting Iran. This came after the Department of Justice submitted a court document related to litigation involving xAI’s data centers.
The court document stated that a version of Grok called the ‘Grok Gov Model' was used in classified military projects. The system had been deployed by Pentagon officials for an AI-based targeting system against Iran.
In the brief, the US Department of Justice argued that the lawsuit “Grok threatens American national, economic, and energy security by seeking to shut off the power supply for artificial intelligence innovation that supports the Department of War's military operations.”
Reportedly, the artificial intelligence system helped process intelligence and identify possible targets within a 96-hour military operation. According to media reports citing different sources, the technology was responsible for identifying almost 2,000 targets and thereby enabling faster operations planning. Officials have not suggested that Grok independently authorized attacks. The technology acted as an aid to decision-making by commanders.
This disclosure shows that generative AI is expanding from its applications in administration and analysis to operations in the military. Earlier this year, xAI was granted permission to use Grok inside the classified US military system.
The revelation is expected to fuel debate over the use of AI in warfare. Critics raised concerns regarding accountability and transparency when using the technology for military purposes and over accidental damage to civilians. Other issues surrounding the evaluation, testing, and regulation of AI before deployment in military settings also exist.
Grok’s reported use in Iran operations signals how rapidly these tools are becoming part of modern military strategy. In light of the increasing adoption of AI by defense agencies, Grok’s role in military actions could be a significant turning point.
Also read: xAI Pushes Grok Beyond Chat with Reusable Workflow Feature