We are a few short months away from the start of a legislative session that will be punctuated with countless debates and hearings on AI legislation. Many of these discussions are dominated by the shiny objects—the deepfakes, autonomous vehicles, robotics, etc. To be clear, these are crucially important topics that demand the utmost scrutiny from policymakers. But we must not lose sight of the less sexy applications which pose some of the greatest opportunities for conservative states to lead and advance human flourishing.
Through responsibly employing AI, we have an unprecedented chance to reduce the size and role of government. And here’s how.
Texas, like a few other states, passed a law last year that established an artificial intelligence advisory council. Experts were appointed to these councils to hear from state agencies on how they are leveraging AI tools currently, as well as opportunities to better take advantage of new and emerging technologies.
While not always the most exhilarating of hearings, representatives from agencies like the Office of the Attorney General, the Department of Transportation, the Workforce Commissions, and so on testify and submit documentation outlining how their agency is reducing fraud, waste, and abuse through AI tools and, ideally, how services are now running more efficiently and cost effectively.
Here are some of the important insights states like Texas are gleaning thus far that can inform future legislation and galvanize public leaders into advancing conservative principles.
Boring is good. Agencies that are leveraging AI to process contracts, invoicing, and other quotidian financial tasks are seeing dramatic reductions in time and money spent. For example, the Texas Department of Transportation has reported a reduction in invoice processing time from three weeks down to 24 seconds, and a 45% reduction in manual labor. Moreover, when agencies have used AI to automate certain components of employee onboarding and offboarding, there has been a significant reduction in process time and manhours devoted to such tasks (an 11-day reduction in process time for one agency!).
AI should inform, not dictate. Texas, like every other state, became overwhelmed with the staggering increase in jobless claims stemming from the pandemic. The Texas Workforce Commission (TWC) was able to stay on top of these claims after rolling out an AI chatbot that walked Texans through how to sign up for unemployment benefits and answered the questions of tens of thousands of Texans daily. Notably, applications like this don’t cut the human out-of-the-loop—rather, they automate the more predictable tasks and enable AI tools to point residents in the right direction. Applying tools like this across government services frees up staff time, gets constituents the information they need quicker, and allows agencies to leverage their talent to innovate rather than maintain the status quo.
AI should increase reach and civic engagement. Local and state governments are employing AI-driven transcription to make hearings more accessible in concert with AI-enabled translation services that ensure larger swaths of residents are reached and engaged. For example, one office in the Texas House of Representatives notes the use of “a spectrum of AI tools to enhance the production process” of an interim committee report, adding a new avenue to make the information from this hearing available to the public while saving staff time.
Local governments and state agencies are experimenting (with mixed success) on the use of GenAI tools to more effectively and palatably present data and information to residents—this can be as simple as asking different AI tools to turn data into charts or as bold as “translating” wonky information into succinct accessible language using prompt engineering.
Humility is key. Texas’ AI advisory council heard from state agencies that are wholly unprepared to start prioritizing using AI to manage different systems. Ultimately, AI runs on data to produce an output. Those state agencies that have terrible data management and governance would be wise to acknowledge the adage “garbage in, garbage out.” In Texas we are learning that slapping AI on top of agencies that do not have their data house in order only makes matters worse from a cybersecurity, privacy, and government efficiency perspective. This is where the ingredients that make AI systems responsible—privacy, transparency, accountability, and human dignity—must be prioritized.
Ultimately, the savings to be had are significant, and leveraging AI should be a focal point for fiscal conservatives throughout the country. One study by Deloitte found that, assuming a modest level of investment and utilization of AI systems in government, between 13 – 15% of employee hours would be saved, saving tens of billions of dollars between federal and state government employees. This is a significant opportunity, and one we should not squander to left-leaning states that are looking to use AI to embed DEI and other woke philosophies further into government.
As states enter their legislative sessions, policymakers should focus on advancing digital codes of ethics and frameworks for state government use of AI that encourages creative applications that save time, money, and constituent headaches.
For all the fanfare about the private sector’s use of AI, conservatives would be remiss to ignore the compelling case for public sector use of AI to rein in the size of government and, yes, align the role of government more closely with what our Founders envisioned—even centuries before the creation of AI.