September 28, 2026 - Personnel Committee Agenda & Packet

www.veronwi.gov City of Verona PERSONNEL COMMITTEE MONDAY, SEPTEMBER 28, 2026 6:15 P.M. VERONA CITY CENTER – CONFERENCE ROOM D122 111 LINCOLN STREET, VERONA, WIS. AGENDA 1. Call to order 2. Roll call 3. Public Comment 4. Approval of minutes from the July 27, 2026 Personnel Committee meeting. 5. Discussion and Possible Action Re: City of Verona Administrative Policy No. 58 - Code of Conduct (Prediction Markets and Event-Based Trading Platforms) 6. Discussion and Possible Action Re: City of Verona Administrative Policy No. 70 - Relating to Generative Artificial Intelligence 7. Adjournment Mara Helmke Chairperson POSTED: Verona City Hall Verona Public Library Miller’s Market City Web Page at www.ci.verona.wi.us Notice is hereby given that one or more members of the Personnel Committee may participate in the meeting by telephone. As required by the Wisconsin Statutes, the City of Verona will provide the public with an effective means to monitor the committee member who is calling in to the meeting by providing speakers in the meeting room. Notice is also hereby given that a quorum of the members of the City Council or other standing committees of the City Council may be present at the meeting of the Personnel Committee to gather information about a subject over which they have decision-making responsibility. The City Council and any other standing committees will not take formal action at this meeting. IF YOU NEED AN INTERPRETER, MATERIALS IN ALTERNATIVE FORMATS, OR OTHER ACCOMMODATION TO ACCESS THE MEETING, PLEASE CONTACT THE CITY CLERK AT 845-6495 AT LEAST 48 HOURS PRECEDING THE MEETING. EVERY REASONABLE EFFORT WILL BE MADE TO ACCOMMODATE YOUR REQUEST. MINUTES – PERSONNEL COMMITTEE DATE: July 27, 2026 LOCATION: Verona City Hall – Large Conference Room 111 Lincoln St., Verona, Wis. 1) Chairperson Mara Helmke called the meeting to order at 6:00 p.m. 2) Roll call – present were: Chairperson Mara Helmke, Alderperson Beth Tucker Long and Alderperson Phil Hoechst. Also present were Mayor Luke Diaz, Alderperson Spencer Harrison, Assistant City Administrator Ann Freiwald, Finance Director Brian Lamers, and City Administrator Jamie Aulik. 3) There was no public comment. 4) Tucker Long moved, Helmke seconded approving the minutes of the May 18, 2026 meeting. The motion passed unanimously. 5) Discussion and possible action re: City of Verona Administrative Policy No. 70 Relating to Generative Artificial Intelligence a) Administrator Aulik explained that the internal AI Workgroup had considered the feedback received at the previous Committee meeting, updated the policy accordingly, and returned the revised policy to the Committee for further review. b) Discussion included whether prompts entered into generative artificial intelligence tools are subject to public records requests. Administrator Aulik explained that Attorney Bryan Kleinmaier determined that prompts are subject to public records requirements and must be retained in accordance with records retention requirements. The Committee also discussed that A.I. companies may have varying retention policies for prompts and that the City’s applicable retention requirements would still apply. The committee asked for more information on how prompts are stored by A.I. companies. c) The Committee discussed the use of Microsoft Copilot, including that it is included as part of the City’s Microsoft Office 365 subscription for City employees and that the City’s current guidance is for employees to use Microsoft Copilot for generative artificial intelligence purposes. d) Alder Tucker Long expressed that she did not want City employees using artificial intelligence. e) There was discussion on which platforms the City would like to use – staff were unsure. There was also discussion on which employees need access to artificial intelligence products. f) The Committee also discussed disclosure when artificial intelligence is involved in generating content and agreed that such involvement should be disclosed. 6) Helmke moved, Tucker Long seconded adjournment. The motion passed unanimously and the meeting adjourned at 6:30 p.m. By: Jamie J. Aulik, City Administrator CITY OF VERONA 111 Lincoln Street Verona, WI 53590-1520 Ann Freiwald, Assistant City Administrator Phone (608) 845- 0919 afreiwald@veronawi.gov MEMORANDUM To: Personnel Committee From: Ann Freiwald, Assistant City Administrator Date: September 23, 2026 Re: Artificial Intelligence Platforms, Governance, and Records Retention Members of the Personnel Committee: • Discussion and Possible Action Re: City of Verona Administrative Policy No. 58 - Code of Conduct (Prediction Markets and Event-Based Trading Platforms) o The growth of prediction markets and event-based trading platforms, where individuals buy, sell, or trade contracts whose value depends on whether specific events occur, raises concerns about how city employees might use non-public information available to them. Ensuring that city employees are not abusing this access to information is an important element of maintaining trust. o The current Code of Conduct Policy covers gambling while at work, the addition of this Prediction Markets section will ensure that the rules surrounding these platforms are clearly stated, extending to beyond just the employee themselves. o The addition to the Code of Conduct Policy strictly prohibits the use of non-public, confidential or insider information by city employees, intermediaries, their family or an organization they are associated with. Any employee who fails to comply with this added Prediction Markets policy is subject to discipline, up to and including discharge. • Discussion and Possible Action Re: City of Verona Administrative Policy No. 70 - Relating to Generative Artificial Intelligence Comparing Three Artificial Intelligences Options (Free, Bundled and Enterprise) As the City of Verona considers the use of Artificial Intelligence (AI) by its staff, the question is which AI option provides the right balance of security, usability, cost, governance, and staff productivity. There are three options when it comes to AI platforms and the pros and cons of each are discussed below. o Option 1: Free AI Platforms (e.g., free ChatGPT, Gemini, Claude free versions) • Pros ▪ No direct licensing cost. ▪ Easy to test and experiment with before making a larger investment. ▪ Staff can quickly learn AI capabilities. ▪ Typically provides access to some of the latest AI models. ▪ Useful for occasional, low-risk tasks such as brainstorming, drafting public communications, or creating meeting agendas. • Cons ▪ Significant data privacy and security concerns. ▪ Limited administrative controls. ▪ No centralized governance or audit trails. ▪ Employees may enter sensitive city information without realizing the risk. ▪ Inconsistent user experience as staff use different tools. ▪ Features, limits, and terms of service can change without notice. ▪ Generally, not integrated with city systems and workflows. ▪ Difficult to support or train staff on multiple free tools. • Best Use Case A short-term pilot or individual experimentation, but generally not recommended as an official citywide solution. o Option 2: Bundled AI Platform (Microsoft Copilot with Microsoft 365) • Pros ▪ Integrates directly with Outlook, Teams, Word, Excel, PowerPoint, and OneDrive. ▪ Employees can work within tools they already use daily. ▪ Enterprise-grade security, compliance, and identity management. ▪ Data remains within the organization's Microsoft environment and permissions structure. ▪ Easier governance through Microsoft 365 administration tools. ▪ Users can ask questions about documents, meetings, emails, and files they already have access to. ▪ Reduces the need to move information between systems. ▪ Easier IT support because it aligns with existing Microsoft investments. ▪ Strong fit for government and municipal operations focused on productivity. • Cons ▪ Licensing costs can be significant. ▪ Highest value is realized when staff are already heavy Microsoft 365 users. ▪ Some employees may not use it enough to justify licensing costs. ▪ Requires training and change management. ▪ AI output quality can vary depending on document organization and data quality. ▪ Organizations may need to review governance, retention, and acceptable use policies. • Best Use Case Knowledge workers who spend substantial time in Outlook, Teams, Word, Excel, report writing, policy development, grants, budgeting, communications, and project management. o Option 3: Stand-Alone AI Platform (Claude, ChatGPT Enterprise, Gemini Enterprise, etc.) • Pros ▪ Often provides strong reasoning, writing, and analysis capabilities. ▪ Can outperform bundled tools for complex drafting and research tasks. ▪ Platform flexibility and choice. ▪ Usually model improvements arrive faster than bundled workplace products. ▪ Useful for policy analysis, strategic planning, report development, and content creation. • Cons ▪ Separate licensing and vendor management. ▪ Additional login and administration requirements. ▪ Less integrated with everyday Microsoft workflows. ▪ Staff may need to manually upload files and transfer content between systems. ▪ Potential duplication of functionality already available through Microsoft. ▪ Requires separate governance and security review. ▪ Multiple AI systems can create user confusion and support challenges. • Best Use Case Smaller groups of power users who regularly perform: • Research • Strategic planning • Policy development • Complex writing • Grant writing • Detailed analysis • Verona-Specific Considerations Strong Arguments for Microsoft Copilot • Most municipal staff already work primarily in Outlook, Teams, Word, and Excel. • Easier governance and records management. • Lower training burden because users work in familiar applications. • Better alignment with existing Microsoft investments. Strong Arguments for a Stand-Alone Platform • If only 10-20 employees need advanced AI capabilities. • If grant writing, policy research, strategic planning, and document analysis are primary use cases. • May provide stronger AI performance for certain analytical and writing tasks. Weak Case for Free Platforms • Difficult to manage from a municipal risk and compliance standpoint. • Potential exposure of confidential, personnel, legal, or citizen information. • Creates inconsistent use and support challenges. o Recommendation Framework A practical approach for the City of Verona could be: • Tier 1: No AI Access Staff whose work is primarily operational, field-based, or transactional. • Tier 2: Microsoft Copilot Licenses Administrative and professional staff who spend much of their day in Microsoft 365. • Tier 3: Advanced AI Users A small group of designated power users (City Administrator's Office, Planning, Finance, HR, Communications, grant writers) who may benefit from either Copilot or a specialized stand-alone solution depending on their work. For the City of Verona this group may be as small as 2 to 5 staff members. o Summary If Verona is looking for the most manageable and lowest-risk organizational solution, Microsoft Copilot is likely the strongest overall option. If the goal is to provide advanced AI capabilities to a small number of high-value users, a stand-alone platform such as Claude may offer better value. Free AI platforms are best viewed as personal experimentation tools rather than an official municipal solution due to governance, retention and security concerns. Record Retention of Microsoft Copilot and Other AI Application Prompts for Public Record Compliance As the City expands its use of Microsoft Copilot and other artificial intelligence (AI) applications, it is important to recognize that user prompts, responses, and related AI-generated content may constitute public records if they are used to conduct public business. Based on guidance from the City Attorney, the City should assume that AI interactions related to official business may be subject to open records requests and applicable records retention requirements. o Microsoft Copilot Microsoft offers varying retention capabilities depending on the Copilot product being used: • Microsoft 365 Copilot operates within the Microsoft 365 environment and may be subject to existing Microsoft retention, eDiscovery, audit, and records management capabilities when interactions are stored within Microsoft 365 services. • Copilot Chat (enterprise versions) may retain chat history and interaction data according to organizational settings and Microsoft retention policies. • Retention periods are configurable through Microsoft Purview retention policies, litigation holds, and records management controls, allowing organizations to preserve content for specified periods. o Other AI Applications Retention capabilities vary significantly among third-party AI platforms: • Some AI tools retain prompts and responses indefinitely unless manually deleted. • Others provide administrative retention settings, export functions, or audit logs. • Consumer-grade AI applications may offer limited retention controls and may not meet government records retention requirements. • Departments should not assume that prompts can be recovered unless a documented retention process exists. o Recommendation The City should establish a formal AI Records Retention Policy that: 1. Defines AI prompts, responses, and generated content as potential government records when used for City business. 2. Requires use of approved AI tools that support retention, auditing, and legal hold capabilities. 3. Aligns AI record retention with existing records schedules and public records laws. 4. Establishes procedures for preserving AI interactions that support decision-making, policy development, public communications, contracts, permits, personnel actions, or other official business. 5. Prohibits the use of unapproved AI services for City business when retention requirements cannot be satisfied. o Conclusion AI-generated content should be treated similarly to emails, documents, and other electronic records when it relates to City business. Because public records obligations may extend to AI prompts and responses, the City should ensure that approved AI platforms provide adequate retention, retrieval, and legal discovery capabilities before they are used for official government purposes. Very respectfully, Ann Freiwald Assistant City Administrator This document may include content generated with AI assistance. City staff are responsible for the final content and accuracy. CITY OF VERONA CODE OF CONDUCT POLICY ADMINISTRATIVE POLICY NO. 58 1. Purpose The purpose of this policy is to clarify guidelines for employee conduct at work. 2. Policy All City employees are expected to meet a standard of conduct appropriate to the reputation of the City. While at work employees are responsible to be aware of and abide by existing rules and regulations. It is also the responsibility of employees to perform their job duties to the best of their abilities and to the standards set forth in their job descriptions, or as otherwise established. Employees, while on City property, while performing their duties throughout the City or representing the City elsewhere and in the conduct of their work, are expected to demonstrate proper regard for the standards of the community and to show respect for the law and for the rights of others. The behavior of employees while not at work or representing the City is for the individual to regulate. However, when illegal and/or inappropriate activity is engaged in which relate to an employee’s job duties and/or which raise doubts as to the employee’s ability to carry out their duties, the activity will be reviewed and appropriate action may be taken a. Expected Conduct Employees are expected to conduct themselves in a positive manner in order to promote the best interests of the City. Examples of appropriate employee conduct include the following (this list is NOT intended to be all-inclusive): 1. Treating all citizens, visitors and co-workers with respect and in a courteous manner; 2. Refraining from conduct that is offensive; 3. Reporting to management suspicious, unethical or illegal conduct by co-workers, citizens or business associates of the City; 4. Cooperating with any City investigation; 5. Complying with all City safety and securing regulations; Title: Code of Conduct Policy Policy Source: Personnel Committee Creation Date: 9/22/2014 Application: All Employees Revision Date: XX/XX/2026 Indexed as: Code of Conduct Policy Total Pages: 5 6. Wearing clothing appropriate for the work being performed; 7. Performing assigned tasks efficiently and in accord with established standards; 8. Reporting to work punctually as scheduled and being at the proper work station, ready for work, at the assigned starting time; 9. Giving proper advance notice whenever unable to work or report on time; 10. Smoking only at times and in places not prohibited by City rules or local ordinances; 11. Maintaining cleanliness and order in the workplace and work areas b. Attendance It is expected that all City employees will arrive at their workplace on time, work all scheduled hours and authorized overtime as required by business necessity, and are dependable with regard to attendance. Excessive tardiness and poor attendance disrupt workflow and customer service and will not be tolerated. Employees who are frequently away from the premises for business reasons are required to inform their supervisor of their whereabouts during working hours. 1. Working overtime Non-exempt employees will not be permitted to work any period of time before or after scheduled starting or quitting times resulting in overtime without approval from their supervisor except in the case of emergencies. 2. Excused Absences Under certain circumstances, an employee’s absence or tardiness may be excused if proper notice is given. Employees that need to be absent or tardy to work for any reason should notify their immediate supervisor as soon as possible upon determination that an absence will occur, unless a verifiable emergency makes it impossible to do so. Issues involving tardiness or extensive absenteeism will be responded to immediately, with consultation with Administration. Unauthorized or excessive absences or tardiness will result in disciplinary action, up to and including discharge. An absence is considered to be unauthorized if the employee has not followed proper notification on procedures or the absence has not been properly approved. 3. Job Abandonment Employees who are absent from work for five (5) consecutive days without giving proper notice to the City will be considered as having abandoned the job. At that time, the City will formally note the termination and advise the employee of the action by certified mail to the employee’s last known address. c. Drugs and Alcohol in the Workplace Employees are prohibited from reporting to work under the influence of illegal drugs and/or alcohol. Illegal drugs and alcoholic beverages are not to be consumed at any time in the workplace. City employees who are impaired while on the job will be dealt with immediately. Employees who are taking any legally prescribed medication, therapeutic drug, or any non-prescription drug which could interfere with their ability to safely and effectively perform their job duties shall notify their supervisor prior to performing work-related duties. The misuse or abuse of legal drugs while performing City business is prohibited. d. Gambling Gambling constitutes a threat to the City’s security, and can interfere with employee productivity and morale. Therefore, any gambling on the part of City employees on City premises or during work time is strictly prohibited. Gambling includes placing, accepting, recording or registering bets, or otherwise carrying on a game of chance for money, property, or any other item of value. This prohibition includes all gambling operations, even if and where such activities do not constitute illegal activity. Employees involved in gambling while at work will be subject to discipline, up to and including discharge. The following list, which is not all inclusive, includes examples of forms of gambling in the workplace that are prohibited: • Any professional or organized gambling activity; • Poker, craps or any card or dice game that involves the exchange of money; • Bingo; • Football, baseball and any other type of “pool”; • Group purchases of lottery tickets. Exceptions to this include office or department sanctioned “pools”, raffles, or a City sponsored event which supports a cause. e. Prediction Markets and Event-Based Trading Platforms Prediction markets and event-based trading platforms refer to any platform which an individual buys, sells or trades contracts, whose value is determined by occurrence or non- occurrence of an event. As these markets and platforms can be influenced by the knowledge of information, it is important to ensure city employees are not taking advantage of their access to non-public or insider information to engage in prediction market or event-based trading platform contracts or betting. City employees, intermediaries, their family or an organization they are associated with, are prohibited from using non-public, confidential or insider information for personal gain, or to advantage themselves in any prediction market or event-based trading platform. • No person shall use the disclosure of confidential, insider or non-public information by a current or former official or employee to participate in prediction markets or event-based contracts, regardless of whether the individual ultimately profits; • No person shall use government resources, databases, investigations, or other nonpublic information or deliberations to obtain an advantage in prediction markets or event-based trading platforms; • No person shall attempt to evade this policy through an intermediary, or assist, encourage or facilitate any other person, in participating in prediction markets or event-based contracts; • Any person aware of violations should report activities to Human Resources. Prediction market or event-based trading platform conduct that interferes with operations or discredits the City will not be tolerated. Employees engaged in prediction markets or event- based trading platforms that violate this code of conduct will be subject to discipline, up to and including discharge. e.f. Insubordination Insubordination, defined as an employee’s failure or refusal to recognize or submit to the authority of a supervisor, or open defiance of authority or resistance to control (i.e. refusing to obey instructions), is strictly prohibited. Insubordination may result in discipline, up to and including, discharge. f.g. Other Prohibited Conduct Any conduct that interferes with operations, discredits the City, or is offensive to citizens or co- workers will not be tolerated. The following are examples of conduct that is strictly prohibited (this list is NOT intended to be all-inclusive). Employees engaged in it will be subject to discipline, up to and including discharge. At management’s discretion, any violation of City policy or any conduct considered inappropriate or unsatisfactory may subject an employee to discipline. 1. Possession of fire arms or other weapons on City property (unless employee is required to carry a firearm or other weapon as a condition of employment); 2. Fighting or assaulting a co-worker or citizen; 3. Threatening or intimidating co-workers, citizens, business associates, or guests; 4. Engaging in any form of sexual or other harassment; 5. Reporting to work under the influence of alcohol, illegal drugs, controlled substances or other narcotics; 6. Disclosing confidential City information; 7. Falsifying or altering any City record or report, such as an employment application, medical reports, production records, time records, expense accounts, absentee reports, or shipping and receiving records; 8. Stealing, destroying, defacing, or misusing City property or another employee’s or citizen’s property; 9. Misusing City communications systems, including electronic mail, computers, internet, and telephones; 10. Refusing to follow management’s instructions concerning a job related matter or being insubordinate; 11. Failing to wear assigned safety equipment or failing to abide by safety rules and policies; 12. Smoking where prohibited by local ordinance or City rules; 13. Using profanity or abusive language; 14. Sleeping on the job without authorization; 15. Gambling on City property; 16. Engaging in pranks or horseplay; 17. Wearing improper attire or having an inappropriate personal appearance; 18. Accepting tips or gifts in the course of work; 19. Engaging in outside employment while on an unpaid leave of absence, while on City premises, and/or using City resources such as copy machines to facilitate non-City related activity; 20. Claiming sick leave under false pretenses; 21. Working unauthorized overtime; 22. Deliberately restricting work output or encouraging another employee to do so; 23. Illegal, immoral, offensive or indecent conduct during the workday or on City property that the employee knows about or should have known CITY OF VERONA ADMINISTRATIVE POLICY NO. 70 GENERATIVE ARTIFICIAL INTELLIGENCE POLICY Title: Generative Artificial Intelligence Policy Policy Source: Personnel Committee Creation Date: XX/XX/2026 Application: Employee Artificial Intelligence Use Revision Date: Indexed as: Generative Artificial Intelligence Policy Total Pages: 5 1. Purpose The purpose of this policy is to set forth requirements City departments will observe when acquiring and using software that meets the definition of “generative artificial intelligence.” While we remain committed to adopting new technologies to aid our mission, when possible, we also understand the risks and limitations of generative AI and want to establish responsible use patterns. Our goal is to balance the City of Verona’s obligation to protect staff and the public from harm with the City’s mission to provide high-quality service. This policy may be reviewed from time to time by the AI Review Task Force and updated as needed to adapt to new technologies and needs. 2. Scope This policy applies to all employees of the City of Verona. 3. Definitions Generative Artificial Intelligence (Generative AI) is a class of computer software and systems, or functionality within systems, that use large language models, algorithms, deep-learning, and machine learning models, and are capable of generating new content, including but not limited to text, code, images, video, and audio, based on patterns and structures of input data. Traditional Artificial Intelligence (Traditional AI) refers to computer systems that use rules, logic, or classical machine learning to analyze data, recognize patterns, and make predictions, without generating new content. An example of this is a spam email filter, which analyses incoming emails and classifies them as “spam” or “not spam” based on learned patterns. While this policy includes principles that apply to AI technologies generally, this policy only applies to Generative AI systems. 4. Artificial Intelligence Principles Principles describe general codes of conduct that represent the City’s values and are aligned with our responsibilities to the residents we serve. These principles serve to guide City employees in their use of both Generative and Traditional AI technology. City employees shall adhere to the principles and requirements outlined in this policy and will be held accountable for compliance with these commitments. 1. Innovation and Sustainability: The City of Verona recognizes the variety of applications for using Generative AI, as well as the possible environmental impacts. The City will make thoughtful and transparent decisions when implementing Generative AI. 2. Transparency and Accountability: The City values transparency and accountability and understands the importance of these values in our use of AI systems. The City will ensure that the development, use, and deployment of AI systems are evaluated for and compliant with all laws and regulations applicable to the City prior to use and will make documentation. Documentation related to the use of AI systems will be available publicly. 3. Validity and Reliability: The City will work to ensure that AI systems perform reliably and consistently under the conditions of expected use, and that ongoing evaluation of system accuracy throughout the development and/or deployment lifecycle is managed, governed, and auditable. 4. Bias and Harm Reduction and Fairness: We acknowledgeThe City acknowledges that AI systems have the potential to perpetuate inequity and bias resulting in unintended harms on Verona residents. The AI Review Task Force will evaluate AI systems through an equity lens, for potential impacts such as discrimination and unintended harm arising from data, human, or algorithmic bias to the extent possible. 5. Privacy Enhancing: The City values data privacy and understands the importance of protecting personal data. WeThe City will work to implement policies and standard operating procedures that reduce privacy risk and are applied to the AI system throughout development, testing, deployment, and use. 6. Explainability and Interpretability: The City understands the importance of leveraging AI systems, models, and outputs that are easily interpreted and explained. WeThe City will work to ensure all AI systems and their models are explainable, and that system outputs are interpretableunderstandable and communicated in clear language, representative of the context for use and deployment. 7. Security and Resiliency: Securing our data, systems, and infrastructure is important to the City. AI systems will be evaluated by the AI Review Task Force for resilience and their ability to maintain confidentiality, integrity, and availability of data and critical City systems, through protection mechanisms to minimize security risks in alignment with governing policy and identified best practices. 5. Policy 1. Acquisition of Generative AI Technology 1.1. City employees may be authorized to use pre-approved Generative AI software tools or they may request a non-standard acquisition of Generative AI software through their department director and the AI Review Task Force. 1.2. The City’s standard for technology acquisition applies to all technology, including free-to-use software or software-as-a-service tools. 1.3. If a technology that has already been approved for use in the City adds or incorporates Generative AI capabilities, no additional approval is required to use those capabilities. However, all other aspects of this policy apply to said use. In consultation with the department director and AI Review Task Force, Verona IT may revoke authorization for a technology that adds AI capabilities, or may restrict the use of those AI capabilities, if, in its judgment, those AI capabilities present risks that cannot be effectively mitigated to comply with this policy or other City policies. 2. Outputs and Human in the Loop 2.1. All outputs generated by AI systems must be reviewed by a human (“(Human in the Loop” or “HITL)”) prior to any official City use. This includes reviewing AI-assisted documents or spot-checking AI-generated analysis. AI-generated content must also be verified for accuracy using reliable sources; if factual information cannot be confirmed, it must not be used for work purposes. While Generative AI may be used to review or edit content, the original source content must be created by a human. 2.2. AI may be used to edit, enhance, clean up, correct, or improve human-created content, including photographs, images, video, text, and audio. However, Generative AI shall not be used to create original content intended for official City use where the primary source material was generated by AI rather than created by a human. Human-created content that has been enhanced through AI-assisted editing remains permissible under this policy. 2.3. Use of Generative AI for software development is prohibited. Employees shall not use Generative AI systems to generate, modify, translate, optimize, or otherwise produce source code, scripts, software applications, automation routines, or executable instructions for City business. 2.1.2.4. 2.2.2.5. All new employees are required to review this policy and train on the proper use of these technologies before use. All applicable staff, such as office staff who may use Generative AI, must complete a refresher course annually to ensure understanding and compliance with this policy. 2.3.2.6. Employees are not required to use Generative AI as part of their job duties. 3. Autonomous Actions and Decision-Making 3.1. Generative AI systems shall not be permitted to autonomously take actions, make decisions, communicate with the public, access City systems, execute transactions, initiate workflows, modify records, or otherwise act on behalf of the City or City employees without direct human review and authorization for each action. 3.2. The use of AI agents or similar technologies that independently act on behalf of a City employee, department, or the City is strictly prohibited. 3.3. City employees shall remain responsible and accountable for all decisions and actions involving AI-assisted systems. 3.4. Attribution, Accountability, and Transparency of Authorship 3.1.4.1. All images and videos modified by Generative AI systems must be attributed to the appropriate Generative AI system. Wherever possible, attributions and citations to the City of Verona should be embedded in the image or video (e.g., via digital watermark). 3.2.4.2. If text generated by an AI system is used substantively in a final product, attribution to the relevant AI system is required. 3.3. If a significant amount of source code generated by an AI system is used in a final software product, or if any amount is used for an important or critical function, attribution to the appropriate AI system is required via comments in the source code and in product footnotes. 3.4.4.3. All attributions should include the name of the AI system used plus an HITL assertion (which should include the department or group who reviewed/edited the content). Example: Some material in this brochure was generated using ChatGPT 4.0 and was reviewed for accuracy by a member of the Department of Human Resources before publication. 3.5.4.4. Departments shall interpret “substantive use” thresholds to be consistent with the principles outlined in this document as well as relevant intellectual property laws. 3.6.4.5. 3.6 Generative AI chatbots may produce content that is plagiarized from its knowledge base, to include copyrighted works. No text generated from a chatbot will be eligible to have City of Verona copyright, trademark, or patent. 4.5. Reducing Bias and Harm 4.1.5.1. Generative AI systems may produce outputs based on stereotypes or use data that is historically biased against protected classes. Each AI system will be evaluated for bias by the AI Review Task Force. 5.6. Data Privacy 5.1.6.1. Employees shall not submit data that is confidential, or that is otherwise not considered to be acceptable to disclose to the public, to Generative AI systems. 5.2.6.2. No City data or records, including inputs or prompts, are to be used for training or parameter-tuning for Generative AI models outside the City’s control. AI technologies that cannot prevent City data or records from contributing to their language models may not be used by City employees. 6.7.Public Records & City Records Management 6.1 Records generated, used, or stored by Generative AI vendors, or solutions may be considered public records. City employees should always proceed as though all records generated, used, or stored by Generative AI vendors or solutions will be subject to disclosure under the Wisconsin Public Records Law. 6.2 All Generative AI solutions and/or vendors approved for City use shall be required to support retrieval and export of all prompts and outputs (either via exposed functionality or through vendor contract assurances). 6.3 City employees who use Generative AI tools are required to maintain, or be able to retrieve upon request, records of inputs, prompts, and outputs in a manner consistent with the City’s rec ords management and public disclosure policies and practices. 6. Exceptions Exceptions must be approved in advance by the employee’s department director and AI Review Task Force. 7. Non-compliance Department directors are responsible for ensuring compliance with and enforcement of this Policy within their departments. Personnel are expected to use official City AI resources appropriately; misuse may result in disciplinary action. All use of AI resources must reflect a high standard of ethical conduct and professional decorum. Failure to meet these standards may result in revocation of access to City Generative AI tools or related features. The input of sensitive or confidential information into AI systems are grounds for termination.