Congressional procurement records can only classify AI tool use where an office logs spending against an identifiable product name, and that structural constraint shapes everything visible about the House's early AI adoption. At least 70 House offices used identifiable AI tools in early 2026, with ChatGPT leading visible spending and Democrats accounting for the larger share of recorded purchases. The actual scope of adoption across the chamber likely runs higher.
The "identifiable" qualifier carries real weight. The 70-office count is a floor, derived from what procurement records can attribute to named platforms. How much additional adoption falls outside that attribution window is not visible in the current data. It is a meaningful gap.
These same offices are actively weighing legislation to regulate commercial AI development. Lawmakers drafting those rules are themselves AI users whose adoption patterns the public record can only partially capture. At least part of that debate will be conducted by the 70-plus House offices already on record, and by others whose use the spending data has not yet registered.