Modernizing Government Responsibly: FOX 5 DC Features Column.us CEO Megan Villanueva to discuss America.gov
By Emmie Atwood
4 Min Read
On September 29, the White House launched America.gov, an AI-powered website intended to help Americans navigate federal information and services through a single portal. That same day, FOX 5 DC asked Megan Villanueva, CEO of Column.us, what the site could mean for citizens and what the administration's growing engagement with AI industry leaders signals about the future of AI in government.
The conversation focused on a challenge Column.us knows well: how to modernize fragmented legacy systems while preserving the safeguards that make them trustworthy. FOX 5 DC host Marina Marraco framed the problem plainly: "I don't know if anyone can really point to any sort of government portal that is efficient, effective, and timely." The question is whether AI can make that experience better and how to modernize responsibly when mistakes have high-stakes consequences.
Modernization, Villanueva argued, can't just mean applying technology for technology's sake. It has to be rooted in outcomes: What problem are we actually solving? Who is the system serving? What are the risks and tradeoffs? And how do we measure whether the outcome is genuinely better — especially for the people we're building it for?
America.gov puts AI at the front door of government
America.gov is designed to become a single digital “front door” for accessing federal information and services. Today, the site uses AI to help people find and understand information published across roughly 29,000 government websites, allowing users to ask questions in their own words instead of first figuring out which agency or website has the answer.
An executive order establishing America.gov directs the General Services Administration, working with the National Design Studio and Office of Management and Budget, to develop it as a “single point of entry” for covered federal services online. The longer-term vision is for users not only to get answers, but eventually to complete government transactions through the platform.
At the launch, President Trump said, “With America.gov, the federal government no longer stands in your way and it stands only at your service.” He added, “We’re simplifying it, we’re glamorizing it, we’re making it what it should be.”
The question: what problem is it solving, and is it actually solving it?
Modernizing access to government information "is promising," Villanueva said, "but it has to be intentional." The test, as she put it, is whether it delivers: "Is this actually in service of the constituents using it? Is it achieving that goal?”
Words like "accessible," "efficient," and "user-friendly" are easy to put in a launch announcement but harder to define in practice. Shaped by her experience leading Column.us, Villanueva offered her perspective on FOX 5 DC:
Accessible should mean more than easier to find and easier for ordinary people to understand and act on.
Efficient should mean a better experience for citizens, not just easier administration for government.
Success shouldn't be measured by traffic or adoption alone, but by whether more people can successfully find and use the public information they need.
The stakes are higher when AI enters government
America.gov runs on generative AI, meaning users may still need to verify important information against the sources it cites, and the site doesn't prominently identify which AI models it uses to answer questions.
Villanueva pointed out that the AI companies themselves are urging caution. "When the actual companies themselves are saying we need to ensure there are safeguards in place as we're developing this over time, we should definitely heed what these tech executives are saying," she said. "When you're actually introducing AI within government, the impact of unintended consequences is much greater.”
She also acknowledged the tension driving the administration's AI push: "How can we stay competitive without sacrificing safety?" And she noted that many tech executives "are actually asking for more government regulation," because they recognize both how powerful the technology is and how much is at stake.
The core concerns are accuracy, transparency and accountability. Underneath all three is a further principle: AI should supplement the human systems and independent checks people rely on, not replace them.
What responsible modernization looks like
At Column.us, we’ve had to think carefully about what it means to modernize legacy, high-stakes legal and governmental systems while preserving reliability, transparency, human support, and trust.
That experience points to four practical commitments when putting AI between citizens and public information:
Build with users. Test the product with the people it is meant to serve, not just the agencies operating it.
Establish clear safeguards. Be transparent about how answers are generated, where they come from and who is accountable when they're wrong.
Measure real outcomes. Test whether the technology is genuinely improving the citizen experience, not just whether people are using it.
Preserve reliable sources. Keep the official records and independent channels people rely on, rather than assuming a better interface solves everything.
As AI becomes part of how Americans interact with their government, the question shouldn't just be how much faster can we make it? It should also be can people trust what they find, and act on it with confidence?
