U.S. Government Website Directs Users to Musk’s Controversial Grok Chatbot

The Trump administration’s dietary guidelines website, realfood.gov, has directed users to Grok, an AI chatbot created by Elon Musk’s company xAI. Grok is the same chatbot that has been linked to generating sexualized deepfakes of women and minors and has faced repeated criticism for producing antisemitic and other harmful content.

When the link first appeared on the government site, users were encouraged to ask Grok nutrition-related questions. People who click through can only use the chatbot briefly before being prompted to create an account. Grok offers paid subscription tiers, and during periods of high demand, users are encouraged to sign up for “higher priority access.” Unless users adjust privacy settings — which are only available to subscribers — their interactions can be used to train the system.

White House officials have defended the decision, saying AI can help people apply dietary guidelines in their daily lives and describing Grok as an approved public-facing tool. Agencies including the Department of Health and Human Services and the Department of Agriculture have also incorporated Grok into internal tasks such as scheduling and communications.

But civil rights and consumer advocacy groups have raised concerns about the government steering the public toward a chatbot with a documented record of generating harmful content. Placing Grok on an official government website lends it credibility and could drive users toward a private platform tied to Musk.

Grok has drawn international scrutiny. Some countries have blocked it, and regulators have opened investigations after incidents involving extremist and sexualized content. Although xAI has said it implemented new safeguards, the problems remain.

The broader concern is simple: a federal government website is directing Americans seeking basic nutrition information to a private chatbot that has a controversial history. For many advocates, that raises questions about oversight, public trust, and whether government endorsement is being used to promote a deeply problematic AI system.