As the world of artificial intelligence (AI) continues to expand, this week showcased government leadership prioritizing funding, regulatory frameworks, ethical considerations, and the deployment of emerging technology pilots across local and federal levels.
Colorado Springs Chatbot
The city of Colorado Springs has introduced “AskCOS,” an AI enabled chatbot designed to enhance constituent engagement by answering questions based on data from the city’s official website. This innovative tool, powered by Citibot and supported by city staff, can respond in 71 languages, providing information on various topics such as wheelchair-accessible parks and city council meeting schedules. The chatbot is integrated into the city’s website and is accessible via text, offering a streamlined way for residents to find information and report issues. City officials will monitor AskCOS to ensure quality responses and to gain insights for improving website content and resident engagement.
Las Vegas Explores AI to Improve Traffic Safety
The Regional Transportation Commission (RTC) of Las Vegas is piloting an innovative program called Advanced Intersection Analytics to enhance traffic safety at high-risk intersections. This initiative leverages AI, predictive analytics, historical data, cameras, and sensors to gather detailed information on traffic patterns and safety issues. The program, in collaboration with local governments and police departments, aims to identify and address violations such as red-light running, lane infractions, and pedestrian non-compliance. Data collected will support enforcement efforts and guide potential engineering changes to improve road safety.
Report Reveals Potential Bias in IRS AI Audit System
A recent report by the Government Accountability Office (GAO) has highlighted potential biases in the IRS’s primary AI tool for flagging audit tax returns. The Dependent Database (DDB) incorporates human inputs, which could introduce unintended racial biases. Despite regular reviews, the IRS has not comprehensively evaluated the DDB’s rules and filters, some of which have not been updated since 2001. This lack of thorough review may contribute to demographic disparities in audit selections, as indicated by a 2023 Stanford University study showing that Black taxpayers are disproportionately targeted for audits. The GAO’s findings urge the IRS to follow an AI accountability framework to address these issues, recommending six actions, all of which the IRS has agreed to implement.
AI in Development to Enhance Government Websites
The General Services Administration (GSA) is developing an open-source tool called “Gov CX Analyzer” to enhance how government websites gather and analyze user interaction data. Announced by GSA Administrator Robin Carnahan during the Workday Federal Forum, this AI-powered tool aims to provide deeper insights into user behavior, moving beyond traditional survey feedback to identify and address friction points on government websites. This initiative aligns with similar efforts by the Office of Management and Budget to improve customer experience performance across federal agency sites.
A recent Statescoop Priorities podcast featuring Indiana and Texas’s Chief Information Officers (CIOs) delves into their states’ strategies for adopting AI technology, particularly emphasizing workforce training and preparedness. The discussion, which echoes concerns raised at the NASCIO 2024 conference, addresses the importance of equipping state workforces with the necessary skills and knowledge to integrate AI into government operations effectively.