TO: Honorable Mayor and Members of the City Council
ATTENTION: Ryan Smoot, City Manager
FROM: Dana Hang, Human Resources and Risk Manager
SUBJECT: Consideration and possible action to adopt an Administrative Policy for Employee Use of Artificial Intelligence (AI).
DATE: June 22, 2026
EXECUTIVE SUMMARY
This action would ratify the City Manager’s implementation of an Administrative Policy and Procedure (AP&P) governing the use of Artificial Intelligence by City employees .
RECOMMENDATION TO CITY COUNCIL
1)Ratify the City Manager’s implementation of AP&P for use of Artificial Intelligence (AI), in a form approved by the City Attorney; or
2)Alternatively, discuss and take other action related to this item.
FISCAL IMPACT
There is no direct fiscal impact associated with adopting the policy. Any future costs related to training or technology tools will be incorporated into departmental budgets.
DISCUSSION
AI tools may assist with improving efficiency, supporting research, streamlining administrative tasks, and enhancing service delivery when used responsibly and with appropriate safeguards. As these tools become more accessible, it is important for the City to establish clear guidelines to ensure responsible, ethical, and secure use of AI tools by employees.
The AP&P governing Employee Use of Artificial Intelligence (AI) provides direction on when and how employees may use AI tools, outlines prohibited uses, and establishes safeguards to protect confidential information, maintain compliance with state and federal law, and preserve public trust. The policy aligns with emerging best practices and supports responsible innovation while protecting the City’s legal, ethical, and cybersecurity obligations.
ATTACHMENTS
Artificial Intelligence (AI) Policy ....................................................................................... 2 Artificial Intelligence (AI) Guidelines ................................................................................ 6 15-C
1 ADMINISTRATIVE POLICIES AND PROCEDURES
APP NO.: 10
Use of Artificial Intelligence (AI)
I.PURPOSE
The City of Bellflower (“City”) is committed to using Artificial Intelligence (“AI”) responsibly, ethically, and transparently in support of City operations and public service. This Administrative Policy and Procedure (“policy”) establishes the foundational principles and requirements governing the acquisition and use of Artificial Intelligence (“AI”) tools by City employees, contractors, and authorized partners. This policy ensures that as AI tools are adopted, they protect the public interest, uphold employee accountability, and safeguard the City’s data and legal obligations.
II.SCOPE
This policy applies to all City employees (full-time, part-time, and temporary), contractors, interns, volunteers, and affiliated partners who use AI tools in any City -related work, regardless of whether the tools are City-procured or used informally. For day-to-day operational guidance, employees must also refer to the City’s AI Guidelines, maintained by the Information Technology (IT) Division and approved by the City Manager.
III.DEFINITIONS
For clarity in this policy, the following definitions apply:
A.AI System. Any model-based tool or platform using artificial intelligence including, without limitation, Large Language Models or “LLMs,” image generators, or classification systems.
B.Generative AI. A subset of AI that produces original content such as text, images, audio, or video in response to user prompts.
C.AI Guidelines. The operational guidance document maintained by the City’s IT Manager that provides specific, updatable direction on approved tools, data handling, and user responsibilities. The AI Guidelines are incorporated into this policy by reference.
IV.CORE PRINCIPLES
All AI use by or on behalf of the City must align with the following principles:
A.Employee Accountability. Employees are responsible for any content, decision, or communication produced with the assistance of AI tools. AI - generated output must be reviewed and verified by the employee before use in furtherance of the City’s business.
2 City of Bellflower – AP&P No. 10
Use of Artificial Intelligence (AI)
B. Human-in-the-Loop Oversight. AI tools serve as assistants, not decision- makers. A human with relevant subject matter expertise as to the final work product must review and approve all AI-generated outputs before any publication or decision is finalized. AI may not be used to make final policy decisions, or decisions on behalf of the City, affecting residents, employees, or other relevant stakeholders without review and approval by a human with relevant subject matter expertise.
C. Privacy and Data Security. Confidential, proprietary, or personal data may only be submitted to vetted and approved AI systems meeting the City’s security and compliance requirements. Submission of such data to unapproved AI tools is prohibited. Employees should be mindful that work product created with the assistance of AI tools may be subject to disclosure obligations under applicable law. Accordingly, utilizing attorney-client privileged materials and work product for any AI purposes including, without limitation, “double-checking” or “verifying” information contained in attorney work product, is prohibited. Any queries regarding such information must be made to the City’s counsel.
D. Transparency and Disclosure. The use of AI tools in City work product must be disclosed as determined by the immediate supervisor overseeing the final work product in the normal course of business. Employees must clearly attribute AI-generated or assisted content in public-facing documents with the following: “This content was produced utilizing the assistance of artificial intelligence tools in accordance with City policy .” AI prompts and outputs may be subject to disclosure under the California Public Records Act and in discovery in the context of litigation or criminal investigations.
E. Ethics and Fairness. AI systems or tools may not be used in ways that produce discriminatory, biased, or inequitable outcomes for residents, employees, or other relevant stakeholders. Employees must remain alert the possibility of AI errors, bias, or inaccurate outputs and take corrective action as needed.
F. Authorized Use Only. Employees may only use AI systems and tools reviewed and authorized by the IT Division. Unauthorized AI systems and tools must not be used for City business. Requests to evaluate new AI systems or tools must be reviewed and authorized by the IT Division.
V. GOVERNANCE AND ROLES
Effective AI governance requires clear accountability across all levels of the organization.
A. City Manager. Provides executive oversight of AI adoption citywide and ensures alignment with the City’s strategic goals and values. 3 City of Bellflower – AP&P No. 10
Use of Artificial Intelligence (AI)
B. IT Division. Reviews and approves AI systems and tools before use; and provides employee training. The IT Manager maintains the AI Guidelines.
C. Department Heads. Ensure employees within their departments are trained on and comply with this policy; communicate AI-related concerns or incidents to IT.
D. All Employees. Comply with this policy; use only authorized AI systems and tools; complete training if required; verify all AI-assisted work and content before use, and; report concerns or incidents to their supervisor or the IT Division.
VI. PROHIBITED USES
The following uses of AI are expressly prohibited. Specific examples and additional context are provided in the AI Guidelines.
A. Entering confidential personnel information, personal data, sensitive financial information, legally privileged communications, or law enforcement data into any AI system or tool not expressly approved for such use.
B. Using AI to make fully automated decisions—without substantive human review—that materially affect residents, employees, or other relevant stakeholders (e.g., hiring, discipline, benefits, or enforcement actions).
C. Generating or distributing misleading, deceptive, or fraudulent content, including deepfakes or impersonations of any person, official, or event.
D. Using AI systems or tools for personal, non-City purposes on City devices or networks.
E. Attempting to bypass City security controls, policies, or legal requirements using AI.
F. Using free, consumer-grade AI tools or services where City data may be used to train public AI models, unless explicitly approved by IT.
VII. AI GUIDELINES
Day-to-day operational guidance—including approved tools, data classification, output verification, public records obligations, disclosure requirements, and the process for requesting new tools—is contained in the City of Bellflower AI Guidelines, maintain ed by the IT Manager and incorporated by reference in this policy. All employees authorized to use AI tools must read and acknowledge the AI Guidelines before use.
VIII. COMPLIANCE AND ENFORCEMENT
4 City of Bellflower – AP&P No. 10
Use of Artificial Intelligence (AI)
A. Non-Compliance. Violations of this policy may be subject to corrective action in accordance with applicable City personnel rules and collective bargaining agreements.
B. Incident Reporting. Employees who become aware of a potential violation, data incident, or AI-related harm must report it immediately to their supervisor and the IT Division. Incidents are managed by IT in coordination with the City Manager’s office.
C. Policy Review. This policy should be reviewed at a minimum every three years, or sooner as desirable. Updates will be communicated to all departments by the IT Division.
APPROVED:
Ryan Smoot, City Manager
APPROVED AS TO FORM: __________
Karl H. Berger, City Attorney
Origination: June 22, 2026
CERTIFICATION AND RECEIPT OF UNDERSTANDING
I certify that I have received a copy of the City of Bellflower Policy for Use of Artificial Intelligence (“AI”). I have reviewed and understand its contents and agree to abide by the principles and requirements in these Administrative Policies & Procedures.
Name:
Title:
Date:
5 June 22, 2026
Administration
AI Guidelines – Responsible Use of Artificial Intelligence Maintained by: IT Manager Version: 1.0 Last Updated: June 22, 2026
1. Purpose. These guidelines provide practical, day-to-day guidance on the responsible use of Artificial Intelligence (“AI”) tools. The guidelines supplement—and do not replace—the Administrative Policy and Procedure No. 10 (“AI Policy”), which governs all AI use by employees (full-time, part-time, and temporary), contractors, interns, volunteers, and affiliated partners in furtherance of City of Bellflower (“City”) business. These guidelines are maintained by the IT Manager and updated periodically as tools and best practices evolve. All persons authorized to use AI tools for City business are required to read and acknowledge these guidelines prior to use.
2. Approved AI Tools.
Only IT-approved tools may be used for City business. Using an unapproved tool for City business violates the AI Policy regardless of how widely used or seemingly safe the tool may be. a. Currently Approved Tools.
1. ChatGPT Business (OpenAI). The City’s primary approved generative AI tool.
Approved for general productivity tasks including drafting, summarizing, editing, and research using non-confidential City data only.
2. Claude (Anthropic). Approved for general productivity tasks including drafting, summarizing, editing, and research using non-confidential City data only.
3. Microsoft Copilot. Approved if licensed by the City, for use within Microsoft 365 applications (e.g., email drafting, meeting summaries, document editing).
Employees must confirm with IT whether this tool is active in your account before use.
4. Embedded AI tools in existing software. Software platforms adopted internally may include embedded AI features. Such features are not automatically approved for use.
Employees who identify an embedded AI feature in any City software must notify the IT Division before use. IT will evaluate the feature and communicate whether it is approved, restricted, or prohibited.
b. Not Approved. Free, consumer-grade AI tools are not approved for City business. This includes, without limitation, the free tier of ChatGPT, Google Gemini, Meta AI, and other similar publicly available tools, or tools licensed with personally purchased licenses. c. Requesting a New AI Tool. To request evaluation of a new AI tool, contact the IT Division and provide the tool name, intended use case, and type of data you would enter. IT will evaluate the tool for security, privacy, vendor terms, and compliance.
3. Data Protection – What You May and May Not Enter.
The AI Policy prohibits entry of confidential and sensitive data into unapproved AI tools. This section provides practical guidance on applying that requirement. a. Permitted – Non-Sensitive Data. The following types of data are generally appropriate to enter approved AI tools:
1. Publicly available information and general research topics.
2. Draft documents, agendas, or reports that do not contain personal or confidential information.
3. General City communications and correspondence not containing resident, employee, or other stakeholder personally identifiable information (“PII”). For purposes of the City’s AI Policy and these guidelines, PII is information that can be used to distinguish or trace an individual’s identity, either alone or when combined with other information that is linked or linkable to a specific individual. Examples include, without limitation, names, addresses, email addresses, telephone numbers, social security numbers, driver’s license numbers, financial account numbers, and employee identification numbers.
6 June 22, 2026
Administration
4. Internal operational content that does not identify individuals or contain sensitive details.
b. Use Caution – Internal / Operational Data. Exercise judgment with internal data. Where possible, anonymize or generalize content before entering it into an AI tool. c. Never Enter – Confidential and Sensitive Data. The following must never be entered into any AI tool, including approved tools, without explicit IT authorization and verified contractual safeguards:
1. Personnel records or HR-related information.
2. Employee, resident or other stakeholder PII.
3. Medical or health information.
4. Law enforcement records or related materials.
5. Legally privileged communications or attorney-client materials.
4. Employee Accountability and Output Verification.
The AI Policy establishes that employees are fully responsible for all AI-assisted or generated work product. The following explains what that means in practice. a. Review Everything. Always review and verify AI output before using it. AI tools can fabricate facts (known as “hallucinations”), produce biased content, or make errors. Never assume output is accurate without independently verifying key information. b. You Sign Off – Not the AI. Never submit AI-generated content as your own professional work without meaningful review and editing. The final work product is yours and carries your professional accountability.
c. No Automated Decisions Affecting People. Do not use AI to make final decisions affecting residents, employees, or other stakeholders (e.g., hiring, discipline, permits, enforcement) without substantive human review and documented approval. See AI Policy. d. Verify Before Publishing. Before finalizing any AI-assisted or generated work product, confirm: the information is factually accurate; the content is free from bias or discriminatory language; it complies with City policies and applicable law, and; it is something you are professionally comfortable putting your name on.
5. Acceptable Use.
The following guidance applies to day-to-day AI use. For the full list of prohibited uses, refer to the AI Policy.
e. Do.
1. Use AI to draft, edit, summarize, translate, or format documents and communications.
2. Use AI to brainstorm ideas, research general topics, or prepare outlines. Note that confidential and/or privileged material, such as confidential information or attorney-client confidential communications, must not be input into AI tools not expressly approved for such use.
3. Verify and edit all AI output before using or distributing it.
4. Use only IT-approved AI tools for City business.
5. Disclose AI use in public-facing documents as described in Section 6 below.
6. Complete required AI training before using approved tools. e. Do Not.
1. Use free or consumer AI tools for City work.
2. Rely on AI output without reviewing and verifying it yourself.
3. Share AI-generated content publicly without review and proper disclosure.
4. Retain AI prompts or outputs beyond their immediate use.
5. Use AI tools for personal, non-City purposes on City devices. 7 June 22, 2026
Administration
6. Disclosure.
a. Labeling and Attribution. When AI tools contribute substantially to the preparation of a public- facing City document, report, or communication, include a brief disclosure statement:
“This content was produced utilizing the assistance of artificial intelligence tools in accordance with City policy.”
Disclosure is not required for minor AI-assisted edits such as grammar checking or light rewording.
7. Monitoring.
a. Monitoring. The IT Division will periodically review AI tool usage to verify compliance with the AI Policy and these guidelines. Department Heads are responsible for monitoring AI use within their teams and reporting concerns to IT.
8. Reporting AI-Related Incidents.
If you become aware of an AI-related incident violating this policy, the Personnel Rules, or applicable law—such as AI-generated content that causes harm, suspected misuse of AI, or a security concern—take the following steps immediately:
a. Stop using the AI tool involved.
b. Notify your supervisor as soon as possible.
c. Contact the IT Division.
d. Preserve all records related to the incident. Do not delete anything.
IT will coordinate the response with the City Manager’s office as appropriate, consistent with the incident reporting requirements of the AI Policy. Retaliation against employees who report incidents is strictly prohibited.
9. Copy of Guidelines Provided to Each Employee. The IT Manager will provide a copy of these guidelines to each employee authorized to use AI tools. All users must, prior to using AI for any City purpose, sign an acknowledgment statement confirming that they have received, read, and understood these guidelines and the AI Policy. Copies of employees’ completed statements will be retained in the individual employee’s personnel file.
Approved: City Manager
Approved: IT Manager
Approved as to form: City Attorney
Date:
8