Project Description

In the digital age, artificial intelligence (AI) stands at the forefront of revolutionizing public administration and service delivery. The City of Lebanon, NH, recognizes the transformative potential of AI to enhance public services, making them more efficient, accessible, and responsive to citizens' needs. However, this potential comes with significant challenges, including privacy concerns, ethical implications, and the paramount need for transparency and public trust. As AI systems increasingly influence various aspects of public life, Lebanon, NH, commits to leading by example in the ethical utilization of AI, grounded in the city's "ADM-143 Use of Artificial Intelligence" policy.

Project Aims

Through this integrated framework, Lebanon, NH, aims to chart a responsible course for AI adoption, balancing innovation with integrity and setting a standard for ethical AI use in public administration. This document outlines the principles guiding the AI auditing and technology review process, underscoring our commitment to using AI as a force for public good and enhancing services while safeguarding our community's rights and freedoms.

Methodology

Our methodological process began by reviewing relevant case studies that focused on existing AI auditing frameworks—namely, those of the Canadian government, Credo, and Microsoft—and AI use policies and guidelines from other jurisdictions—Boston, California, Georgia, San Jose, Singapore, and Utah. Based on the findings of our case study research, we developed an AI Auditing Framework that prioritizes privacy, security, equity, transparency, reliability, and sustainability. We then generated an AI Assessment Worksheet that serves as the tool for implementing the AI Auditing Framework. Using our AI Assessment Worksheet, we completed a use case analysis for Policereports.ai, a software currently under review.

Project Deliverables

Our project delivered three key components to enhance Lebanon's AI governance:

• AI Auditing Framework: This policy will establish the risk assessment, evaluation criteria, and audit procedures for the City of Lebanon's AI auditing framework. AI audits will be conducted by the City's Technology Review Committee (TRC) and ensure the AI software selection and usage process align with the City's values and responsibility to residents. Key principles will guide the TRC throughout the auditing process to ensure that AI technologies are used ethically and transparently.

• AI Software Review Worksheet: The AI Software Review Worksheet provides a framework for the City of Lebanon to assess the risk of AI applications adopted by the city and ensure that these applications adhere to the guidelines of the AI auditing framework.

• The AI Assessment Questionnaire poses questions to vendors to provide Lebanon with digestible information about important service features and demonstrate consideration given to risk, system integration, and responsible use concerns.

Project impact and future work

In conclusion, the AI Auditing Framework for Lebanon, NH, represents a significant step forward in ensuring that the deployment and use of artificial intelligence within city operations are conducted in a manner that is ethical, transparent, and aligned with the community's values. This document, rooted in a deep understanding of the dual nature of AI's impact, sets a benchmark for responsible AI utilization, emphasizing the importance of safeguarding individual privacy, ensuring equity, and maintaining robust oversight. The detailed strategies and proposed changes to existing policies underscore the city's commitment to leading by example in the ethical use of AI technologies.

Moving forward, it is recommended that the City of Lebanon, NH, implement the AI Auditing Framework through a phased approach, beginning with the establishment of the AI Auditing Sub-Committee and an integration of AI risk assessment models into existing technology review processes. Training for TRC members on the nuances of AI technologies and the creation of a public engagement mechanism to gather and incorporate community feedback on AI use are crucial steps that will enhance the framework's effectiveness and relevance. Additionally, updating city policies to reflect the incorporation of AI technologies, as outlined in the proposed changes, will solidify the city's stance on responsible AI use.

Contributors