Tech Policy Press: US Congress Moves to Take Advantage of AI
BY AUBREY WILSON (as published in Tech Policy Press)
On January 30, Members of the US House of Representatives plan to explore a new perspective in the legislature’s discussions on artificial intelligence (AI). The Committee on House Administration’s hearing on “Artificial Intelligence: Innovations within the Legislative Branch'' will focus on Congress’s approach to navigating the emergence of these new technologies in its own work and processes, rather than on what it should do to regulate them. It comes on the heels of a Senate Rules and Administration Committee hearing last week in which witnesses from the Library of Congress, Government Publishing Office, and Smithsonian Institute engaged in an informative discussion about how these Legislative branch support agencies are beginning to respond to both the benefits and risks posed by AI, from collection access to copyright implications.
These forward-looking conversations focused on the US government’s own strategic adoption of technology are a marked shift in Congress’ approach from decades past. While Congress grapples with important questions concerning future legislation to regulate AI — particularly generative AI — quiet, bipartisan work is proceeding steadily to ensure that Congress itself takes a responsible approach on internal adoption of these technologies for day-to-day government operations.
In fact, the House of Representatives is currently among the world’s leading legislatures in setting policies and encouraging responsible experimentation with generative AI. Since the introduction of ChatGPT in November 2022, the House and Senate have provided institutional guidance for internal use without curtailing experimentation and innovation. Early last year, the House established an AI working group managed by the Chief Administrative Office (CAO). The Senate took a similar approach, launching its own working group in November.
In September, the Committee on House Administration’s Modernization Subcommittee began requiring monthly AI transparency reports from House support agencies, listing potential AI use cases and updating on pilot programs, the creation of advisory boards, and other initiatives. These reports, which are summarized and posted publicly in the form of AI Flash Reports on the Subcommittee's website, are the first of their kind to be regularly reported by a representative body. The four reports released in 2023 showcase AI exploration across all participating entities — including the establishment of an AI advisory committee at the Government Publishing Office (GPO) , experimentation with machine learning capabilities to aid information access at the Library of Congress’ “LC Labs,” and updates on the AI-enabled Comparative Print Suite by the Office of the Clerk.
These innovative and proactive approaches are a breath of fresh air from a legislative body not commonly recognized for quick adoption of modern practices. The cringe-worthy moments of 2018’s big-tech hearings still loom large in the public’s view of the tech savviness of their elected officials.
However, the House has shown particular deftness in its response to generative AI. This is largely due to the groundwork laid by the Select Committee on the Modernization of Congress, an extremely effective bipartisan group of 19 Members whose 202 recommendations for improvements to the House promoted more agility and accountability. The Select Committee’s legacy continues as the Modernization Subcommittee continues its work and dozens of the recommendations have been implemented. This ongoing forum for Members to raise potential improvements or upgrades has pushed an institution otherwise reluctant to change to respond to new ideas, try new approaches, and adopt new technologies.
The pandemic also accelerated modernization efforts, compelling the House to quickly adopt new technologies and policies to continue operations. In a matter of months (March through June 2020, to be exact) the House had implemented more IT modernization than it had in the decade prior, according to experts.
These upgrades included the House Clerk’s launch of the eHopper, which finally allowed for the online submission of legislative documents for official consideration. House and Senate cybersecurity offices authorized use of web conferencing platforms Zoom, Webex, and Microsoft Teams — virtual meetings and remote work were previously stark anomalies in the halls of Congress — and many staff were issued laptops with remote access keys for the first time. The Secretary of the Senate developed and launched Quill, a customized e-signature solution, which was quickly adopted by the House as well.
These notable successes improved Congress’s confidence in its own ability to explore and apply new technologies. The chambers proved to themselves and their Members that they were able to adapt quickly and mitigate risk in the process. Today, these operational improvements are still in effect — with efficiency gains that have continued long past the pandemic.
The House not only benefited from the reforms promulgated by the Select Committee on the Modernization of Congress and its rapid adaptation during the pandemic, but emerged as the leader between the two legislative bodies in assessing and adopting new technologies. Although modernization efforts have not taken as strong of a root in the Senate, its recent AI guidelines underscore the potential that now exists for the chambers to advance together.
Rarely is Congress not considered “behind” in understanding technology, let alone adopting it. AI can now be the next catalyst that the government needs to prioritize rapid innovation in its service delivery, customer experience, and day-to-day functions. It is clear the spark is already there: our institutions are in the race to safely harness AI, rather than back at the start line.
