Congressional lawyers are drowning in AI-drafted bills as Capitol Hill experiments with new tools like the Byrd Bot, leaving counsel to clean up the chaos and guard against costly mistakes.
If you were thinking of a steady job as a congressional lawyer, brace yourself: offices are getting flooded with machine-generated legislative text. Staffers and counsel are spending hours fixing drafts that look plausible on the surface but collapse under legal scrutiny.
The speed of generative AI means anyone with an idea can produce stacks of legislative prose in minutes, and the House Office of Legislative Counsel is the one left to sort through it. That extra volume is stretching resources thin and creating fresh risks for lawmakers who assume the drafts are ready to go.
Congressional lawyers are struggling to manage a flood of legislative proposals generated using artificial intelligence tools, many of them riddled with errors and shoddy language.
The strain on the House Office of Legislative Counsel is increasing as more congressional offices and outside groups rely on generative AI to draft legislation, according to interviews with eight current and former officials who work or have worked with the office.
They said AI’s ability to quickly generate mountains of text is forcing the OLC’s lawyers to spend significantly more time reviewing and rewriting the proposals. And in many cases, the drafts are plagued by mistakes that could ripple well beyond Washington — tying up the passage of legislation and risking lawsuits over erroneously cited statutes or incorrect legal definitions.
AI’s arrival on Capitol Hill means that the rapidly advancing technology — already threatening to disrupt software, warfare, Hollywood and myriad other industries — could soon shape laws that affect millions of Americans, in unpredictable ways.
“AI is good for a lot of stuff,” said Daniel Schuman, executive director of the nonprofit American Governance Institute, which helps modernize government technology. “But it’s not capable of drafting legislation that you would want to enact into law.”
Staffers and outside groups are “going to Claude or ChatGPT to draft the legislative language itself,” said one person who has advised office staff on the technology, who like several other people quoted in this story were granted anonymity because they were not authorized to speak on the record.
Given the errors AI tools generate, the OLC spends “more time trying to fix AI-drafted legislation than it would take for them to draft it from scratch,” the person said.
On the Republican side, aides have been experimenting with a tool called the Byrd Bot, designed to help draft language that survives the Byrd Rule in the Senate. The idea is straightforward: use automation to test whether provisions are likely to be tossed from reconciliation bills and to produce cleaner drafts up front.
A new AI tool, called the Byrd Bot, was developed by the conservative Republican Study Committee to help House Republicans draft legislation that follows the Byrd Rule, a Senate provision that allows senators to remove “extraneous” provisions from budget reconciliation bills. The Byrd Bot is used for testing new ideas against the Byrd Rule and for generating suggested changes to legislation text. The hope, they say, is to avoid the frustrating process where House provisions are jettisoned at the last minute due to Senate rules.
https://x.com/AniruddhaSalve_/status/2081342155751399772
The Byrd Bot is the latest example of AI being used in the corridors of power, and the first AI tool to be used in the legislative drafting process. The House previously introduced the Comparative Print Suite in 2022, an AI tool used to compare different versions of legislation and to suggest potential amendments. Some members of Congress have also delivered speeches written and voiced by AI.
It’s important to note that AI tools are not a partisan toy; both sides of the aisle have tried machine drafts. Still, the Republican push to build guardrails into the process is deliberate—create tech that respects Senate rules and reduces last-minute chaos. That approach treats AI as a tool to help professionals, not replace them.
Critics warn of real harms if lawmakers over-rely on AI without oversight. “Could introduce errors, create unintended legal consequences, or result in language that does not accurately reflect a member’s policy intent,” said Democrat Rep. Joe Morelle (NY-25), echoing a common concern on Capitol Hill.
Experts outside government are watching nervously. Kay Firth-Butterfield, CEO of Good Tech Advisory, put it bluntly: “We are leaping ahead into a world where AI creates our laws without actually agreeing that AI should create our laws. We’re simply passive passengers on this.” That nails the risk: technology moving faster than the rules and habits that make lawmaking reliable.
Practical politics matters here. Republicans using targeted tools like the Byrd Bot are trying to thread a needle—use automation to meet procedural tests while keeping lawyers in the loop to vet legal accuracy. That’s a different mindset than handing raw AI output straight to the floor.
Whatever the approach, Congress will need clearer norms and real human review if AI stays part of the legislative process. Lawmakers can exploit clever tools to make government more efficient, but they also have to own the legal consequences when drafts go bad. For now, the cleanup falls to the counsel, and their workload is only growing.
