What's Happening?
Vermont is actively implementing artificial intelligence (AI) across various government functions, transitioning from an ethical framework development phase to an enablement phase. The state appointed Josiah Raiche as its first director of AI in 2022,
later elevating him to chief data and AI officer in 2024, integrating AI governance with its broader data strategy. A key application involves training developers to use AI tools to identify and harden vulnerabilities across approximately 1,100 state applications, starting with cybersecurity. These AI tools can scan source-code segments, model attack paths, generate Python checks, identify vulnerabilities, and support remediation. Additionally, AI is being used to automate routine tasks, such as ballot proofing for the Secretary of State's office, significantly reducing manual effort and improving efficiency under tight deadlines.
Why It's Important?
Vermont's proactive adoption of AI in government is significant as it demonstrates a commitment to modernizing public services and enhancing cybersecurity in an era of increasing digital threats. By using AI to identify and fix vulnerabilities in its applications, the state is strengthening its defenses against cyberattacks, protecting sensitive citizen data, and ensuring the continuity of essential government operations. The automation of routine tasks, such as ballot proofing, frees up human resources to focus on more complex and strategic work, potentially leading to greater efficiency and cost savings for taxpayers. This approach also positions Vermont as a leader in responsible AI implementation, having established an ethical framework early on, which could serve as a model for other states grappling with the integration of AI into public administration.
What's Next?
Vermont plans to continue training more developers in AI tools, with an additional 50 expected to participate, building on the approximately 100 already trained. The goal is to further embed AI into daily operations, allowing developers to maintain and build applications more effectively. The state will also continue its 'pathfinder' model, encouraging AI adoption within various agencies by training trusted employees to develop practical AI applications for their teams. This bottom-up approach aims to identify and implement high-return use cases that make employees' work more efficient and improve constituent services. Future developments may include expanding AI applications to other areas of government, guided by the state's ethical framework and a focus on human oversight.
Beyond the Headlines
Vermont's strategy of viewing AI as a 'power tool' operated by humans, rather than an autonomous agent, highlights a crucial philosophical approach to technological integration in government. This perspective emphasizes human control and accountability, addressing common concerns about AI's potential for unintended consequences or job displacement. By focusing on automating 'drudgery,' the state aims to empower public servants to engage in more creative and mission-oriented work, potentially boosting morale and job satisfaction. This model could influence how other governmental bodies approach AI, promoting a balanced view that leverages AI's capabilities while maintaining human agency and ethical considerations. The success of Vermont's initiatives could also foster public trust in AI applications within government, demonstrating tangible benefits without compromising core values.
