The US Army Ran Out of AI Credits. Here Is What That Tells Us.
The Army promised its workers unlimited AI access. Six weeks later, the credits were gone. Now soldiers and civilian staff face new limits, and one worker says the tools were unreliable anyway.
Key points
- The US Army's annual allotment of 100 million AI tokens (credits for generating text or images) ran out by mid-June 2026, roughly six weeks after the Army's chief information officer announced unlimited access.
- The Army uses a platform called Ask Sage, which lets staff run several AI models including Google's Gemini, Meta's Llama, and OpenAI's ChatGPT.
- Each employee was given at least 200,000 tokens per month, and the system automatically topped that up if they ran out.
- It is unclear whether the Army's token pool will be renewed after 1 October 2026.
- Other large organisations, including Meta and Uber, have hit similar walls after pushing heavy AI adoption.
The US Army spent the first half of 2026 telling its workers to use AI as much as possible. In May, the Army's chief information officer announced that access to the service would be unlimited. By mid-June, the entire year's supply of credits had gone.
An internal email, first reported by Wired AI, confirmed the shortfall and warned staff at the Army's Combat Capabilities Development Command to slow down. "The Army CIO pool was exhausted of tokens," it read.
Tokens are the basic unit of work an AI model performs. Every word it reads or writes uses a few of them. For the Army's system, one token covers about 3.7 characters, roughly one short word. The Army had bought 100 million of them in an annual subscription. The entire stockpile was gone in under two months.
What does this mean for Army staff and their work?
For now, limits are back in place, but there is no guarantee the pool will be refilled when the government's financial year ends on 1 October. Staff who relied on the tools for everyday tasks face uncertainty about whether those tools will still be available.
One Army employee, who spoke to Wired AI without using their name because they were not cleared to talk to journalists, said they did not find the AI tools reliable. One model told them it had finished a task it had not actually done. "An unthinking application and use is not going to result in an effective, efficient, and trustworthy rollout," they said.
The Army uses Ask Sage, a platform that connects to several large language models, the technology behind AI chatbots like ChatGPT, as well as Google's Gemini and Meta's Llama. The service is approved for handling controlled but unclassified government information, such as rewriting job descriptions.
The Army and the Department of Defense did not reply to requests for comment. Ask Sage also did not respond.
The Army is not alone in this. Meta encouraged its own engineers to maximise their AI use, then quietly removed the leaderboard tracking who used the most. Uber reportedly burned through a full year of AI credits in just four months.
The pattern is consistent. Organisations push rapid AI adoption, staff take the invitation seriously, and the bills arrive faster than anyone planned.
Whether the Army's experience prompts a more careful rollout, or simply a bigger budget request, is an open question.
