GovFacts articles are produced by a system we call the GovFacts Engine: software that turns a government topic into a clear, sourced and maintained explainer. It has produced every new article since January 29, 2026.
Articles published before the Engine are in the process of being rebuilt to the current standard.
Key Steps
Choosing topics. We look at the questions people bring to government: what they search for, what they ask us, and what is in the news. A person decides which topics are queued. We welcome reader suggestions.
Planning. For each topic, the Engine lists the questions a reader will ask: what it is, who decides, who qualifies, what it costs, how to apply, what is changing. On a contested topic it also lists each position that identifiable advocates hold.
Gathering sources. The Engine finds and retrieves the documents that answer those questions: government and agency documents, intergovernmental bodies, research and academic work, organizations and advocacy groups, industry and standards bodies, and news and analysis. Each position on a contested topic is sourced from its own advocates. A typical article reviews 20 to 30 documents.
Writing from the sources. The Engine writes only from passages it found in those documents. Each factual sentence is tied to the exact passage it came from and linked to that source. Quotation marks go only around a source’s exact words.
Independent checks. Separate AI readers, which did not write the article, then check it:
- A review of the whole draft for facts, logic, fairness, readability and style, in that order of importance.
- A sentence-by-sentence check that each factual sentence stays inside what its passage supports.
- A final read of the finished article.
- A legal review.
Publishing. An article that does not pass its checks is not published. A published article appears under the GovFacts name with its sources shown.
Keeping pages current. Government does not hold still. Articles built by the Engine are re-checked on a schedule, and each shows when it was last checked and when its next check is due. When a reader reports a problem, we re-check the article against its sources.
The AI Behind It
The Engine runs on frontier AI models. AI Use and Oversight says which models we use, what people do, and where the limits are.
Delivering on Our Priorities
Getting the facts right is the first job, but it isn’t the only one. An article can be accurate and still be confusing, one-sided, or carelessly built. So beyond fact-checking, every article the Engine builds goes through a battery of checks. Some judge the quality of the writing, and others make sure our own automated editing didn’t break anything along the way.
As an article is prepared, here are some of the dimensions that matter to us:
- Fairness. Are competing positions shown and presented in their best light?
- Plain language. Is it written in words an ordinary person would actually use, not bureaucratic or academic jargon?
- Substance. Does each section say something concrete (real numbers, names, rules, examples, etc.) or does it stay vague?
- Logic. Does each conclusion actually follow from what came before it?
Checks are ranked by how much they matter. Getting a fact right counts for more than polishing a sentence, so problems are prioritized in order: facts first, then logic, then fairness, then readability, then style. A low-priority change, like smoothing some wording, is never allowed to alter a fact or break the logic of an argument.
Responsible AI and Our Limits
GovFacts was recognized as a Webby Honoree for Responsible AI (2026). This is recognition we take as a standard to keep meeting, not a finish line.
We show our sources, check claims against them before publishing, keep pages under review, and correct errors when we learn of them. Even so, AI makes mistakes. For any decision that affects you directly, confirm the details with the official government source. AI Use and Oversight sets out the limits.