
Reuters, citing the opinions of seven federal employees, three contract experts, and a review of U.S. government AI inventory documents, found that Grok of xAI, the artificial intelligence company that Musk merged with SpaceX in early February, is not frequently mentioned in deployment systems across federal agencies, with the exception of the U.S. Department of Defense.
A 2025 consolidated inventory from federal agencies revealed over 400 publicly available examples of AI use in government that specifically named vendors. Of these, only three examples involved the use of xAI or Grok. In contrast, 234 entries mentioned OpenAI technologies, including ChatGPT, Codex, and Microsoft Copilot; 33 examples involved Gemini or other Google products; and 26 examples involved Anthropic’s Claude – a product that the Trump administration later ordered to be discontinued.
According to the list, Grok was supplied to US federal agencies for eight months at a cost of 42 cents per agency. This near-zero rate is also used by many other AI companies – a typical strategy to attract clients, which can then lead to higher-priced contracts later.
The list was compiled by the U.S. Office of Management and Budget (OMB). Companies like xAI, Google, and OpenAI will sign contracts through the U.S. General Services Administration (GSA).
xAI, OMB, and GSA declined to comment.

Grok AI application logo. Photo: Lưu Quý
Dựa theo Thương nhân trong cuộc , OMB data raises questions about whether Grok can gain AI market share from leading products like Claude or ChatGPT, especially as SpaceX prepares for what is expected to be the biggest IPO in history, with ambitions to value the company at $1.75 trillion.
Last month, xAI also failed in its bid to deploy Grok for the U.S. Department of Veterans Affairs because it “did not meet the department’s requirements,” according to a source close to the matter. The U.S. Department of Veterans Affairs declined to comment.
In the US business environment, the situation is no better. According to a report published late last year by web traffic monitoring company Netskope, Grok “has not gained significant traction,” with only 2 out of 1,000 users, down from a peak of 5 out of 1,000 users previously. “Even among existing users, the usage time is very short. Grok will not be able to become popular in the US business world,” commented Ray Canzanese, CEO of Netskope.
Previous statistics also show Grok losing ground in the AI race, according to WSJ . For example, according to analytics firm AppMagic, app downloads totaled only about 8.3 million in April, a 58.5% decrease from the record high of over 20 million in January. A survey by research firm Recon Analytics of over 260,000 US AI users showed that Grok’s pay-per-view rate remained virtually unchanged in the first four months of the year, at 0.174%, while ChatGPT reached 6%.
In the field of programming assistants, where competition among AI companies is fierce, Grok is also considered to be lagging behind. According to Erik Bradley, research director at market analysis firm Enterprise Technology Research (ETR), Grok has seen virtually no growth in organizations and businesses. Meanwhile, the use of Claude and Gemini has skyrocketed. A survey of 500 people in March by ETR found that 48% of respondents said their businesses are using or will use Claude, up from 21% the previous year; 40% are using Gemini, up from 27%; but only 7% are using or plan to use Grok, up from 4%.
However, the OMB’s inventory documents do not include the Pentagon, which has a $200 million agreement with xAI. In May, xAI became one of seven companies deployed on the U.S. Department of Defense’s classified networks.
Bảo Lâm compilation
- Elon Musk’s Grok is losing ground in the AI race.
- Grok’s AI continues to be found creating pornographic images.
- Former Grok engineer 3: Focus more on math than opening AI classes.
- Grok ‘created three million sensitive images in 11 days’.
- Grok stops creating sensitive images of real people.

Trả lời
Bạn phải đăng nhập để gửi phản hồi.