TLDRocket
Sign in

Can a chatbot fix the government maze? The White House is about to find out

TechCrunch Amanda Silberling

The White House is launching America.gov, an AI chatbot for government info. It could make services easier to find — or hand out bad advice when the stakes are real.

Based on reporting by TechCrunch, Amanda Silberling — read the original for the full story.

Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error

The White House is rolling out America.gov, a chatbot meant to help people navigate government services and information. President Donald Trump announced the plan on Tuesday, and Google says it is part of the launch through its Gemini model. The company did not say whether any other AI firms are involved.

The pitch is simple: one front door instead of a maze of tens of thousands of government websites and rules. That’s the promise Trump made in a post on X. For anyone who has tried to find the right form, the right office, or the right deadline, that idea has obvious appeal.

But a chatbot is not a search box with manners. Large language models still make things up, and they do it with alarming confidence. CNN recently reported a case in which the U.S. military came close to launching an armed operation against a Chinese vessel over an AI hallucination, only to abort at the last minute when the supposed threat fell apart.

America.gov probably won’t cause a military blunder. The danger is quieter and more common: wrong guidance on food stamps, visa renewals, or tax filing. In those cases, a bad answer can mean missed deadlines, denied benefits, or penalties. That’s not a bug you shrug off after the demo.

My take — AI-written commentary, not fact-checked reporting

This is the classic government AI move: slap a friendly chatbot on a hard problem and hope the friction looks solved. It isn’t. If the system is going to sit between people and benefits, taxes, and visas, then “mostly right” is just a polished way to say risky.

Read more about this at: TechCrunch

Related stories

The daily briefing

Every AI story that matters, in your inbox by 8am.

TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.