AI Policy

Responsible AI: Policies and Guidelines for County Employee Use of Generative AI Tools

Our updated guidelines enable employees to utilize AI tools effectively, while ensuring that their use aligns to County values. Through implementing clear guardrails and training, we can leverage AI to enhance our operations without compromising our commitment to accountability and public trust. Below are the areas the County policy will focus on.  

These policies will evolve as these emerging technologies advance. Adhering to these policies and guidelines ensure Miami-Dade County employees can responsibly harness the power of AI to enhance public services, improve efficiency, and create a more responsive government for County residents.

Policy: Adhere strictly to data privacy standards and security protocols as managed by the ITD’s Cybersecurity division.

Guideline: Never input sensitive County data, personal information, or confidential materials into public AI tools such as ChatGPT, Perplexity.ai, etc. Protect sensitive information and prevent unauthorized access.

Do not post sensitive information that could put County data in jeopardy.
Policy: Ensure that all AI-generated content undergoes thorough human review and validation before finalization, dissemination, or implementation.

Guideline: Always verify, fact-check, and critically evaluate any content generated by AI tools before using it in official communications, documents, decision-making processes, or public services. AI-generated content should be considered a starting point and not a substitute for professional judgment, expertise, or independent verification.
Policy: Complete all mandatory employee training programs related to AI and provide constructive feedback to enhance the effectiveness of these programs.

Guideline: Actively participate in County-provided training sessions, workshops, and online modules on responsible AI use, data privacy, cybersecurity best practices, and ethical considerations. Share your insights, suggestions, and concerns through Countywide AI surveys and tool evaluations to contribute to the ongoing improvement of our AI strategy.
Policy: Stay consistently updated on the capabilities, limitations, evolving functionalities, and potential risks associated with the AI tools you use.

Guideline: Engage in ongoing education, self-directed learning, and professional development activities to stay abreast of the latest AI advancements, emerging trends, and innovative applications in government services. Participate in relevant workshops, conferences, webinars, and online communities to expand your knowledge and expertise.

1. Data as a Strategic Asset: The County recognizes data as a valuable asset that can drive informed decision-making, improve operations, and enhance public services.
2. Data Quality: The County is committed to maintaining high data quality standards, ensuring accuracy, completeness, consistency, and timeliness of data.
3. Data Security and Privacy: The County prioritizes the protection of data confidentiality, integrity, and availability through robust security measures and compliance with applicable privacy regulations.
4. Data Accessibility and Sharing: Data should be accessible only to authorized users for legitimate purposes, while ensuring appropriate controls and restrictions to protect sensitive information.
5. Data Accountability: Data owners and stewards are responsible for the management and governance of their respective data assets.
6. Data Governance Framework: A comprehensive data governance framework will be established to oversee the implementation and enforcement of this policy.
- Data Governance Council: An executive-level Data Governance Workforce will be formed to provide strategic direction and oversight for data governance activities.
- Data Stewardship Roles: Data owners and stewards will be assigned to manage specific data assets, ensuring their quality, security, and appropriate use.
- Data Standards and Policies: Consistent data standards, definitions, and policies will be developed and enforced to ensure data consistency and interoperability. Following best practices in classifying data, a comprehensive Data Classification Framework will be developed to identify and categorize data based on its sensitivity, confidentiality, and criticality. Exemplary levels are included below:
  - Level 1 (Public Data): Information intended for public access that poses minimal risk if disclosed.
    - Examples: Published reports, public records, website content, press releases.
  - Level 2 (Internal Data): Non-sensitive information accessible to internal County employees but not intended for public distribution.
    - Examples: Drafts of policies, internal procedural documents, non-public building layouts.
  - Level 3 (Confidential Data): Information restricted to authorized personnel only, where unauthorized disclosure may cause material harm.
    - Examples: Employee records, non-public financial information, and data with personal identifiable information (PII).
  - Level 4 (Highly Confidential Data): Highly sensitive data requiring strict access controls due to its potential to cause significant harm if disclosed.
    - Examples: Social Security Numbers, financial account details, health/medical information, infrastructure/secure facility drawings/building plans, and encryption keys.
- Data Management Processes: Effective data management processes will be implemented for data collection, storage, retrieval, analysis, and disposal leveraging Data Classification Framework. Departments will classify their data assets and ensure appropriate safeguards, access controls, and handling processes are implemented in accordance with classification level.
- Data Security and Privacy Controls: Robust security measures will be implemented to protect data from unauthorized access, use, disclosure, disruption, modification, or destruction.
- Data Quality Management: Data quality assurance processes will be established to monitor and improve data accuracy, completeness, consistency, and timeliness.
- Data Sharing and Collaboration: Guidelines for data sharing and collaboration will be developed to facilitate information exchange while protecting sensitive data.
- Data Quality: Ensure data accuracy, completeness, and consistency.
- Data Bias: Mitigate bias in data and algorithms.
- Ethical Considerations: Adhere to ethical guidelines for data use and AI development.
- Transparency and Accountability: Be transparent about data sharing practices and AI models.
- Continuous Improvement: Regularly evaluate and improve data sharing processes and analytics capabilities.
Abstract Text for AI Implementation: Learn more about Miami-Dade County’s approach to ethical and responsible AI implementation, featuring use cases and toolkits designed to empower our workforce and residents, and best practices.

Resource Toolbox

Access guides and educational resources that will be updated regularly.

Interest Categories

Online Services

News & Social Media

Create a new miamidade.gov account

Elected Officials

County Agencies

Public Records & Open Data

Employee Information

My Employee Portal

Interest Categories

Online Services

News & Social Media

Create a new miamidade.gov account

Elected Officials

County Agencies

Public Records & Open Data

Employee Information

My Employee Portal

AI Policy

Responsible AI: Policies and Guidelines for County Employee Use of Generative AI Tools

Authorized Tools

Collaboration

Data Protection

Content Verification

Ethical Use

Respect Copyrights

Transparency

County AI Trainings and Surveys

Continuous Learning

Reporting

Data Governance Principles

Data Governance Framework

Key Considerations for Meaningful Analytics and Trusted Data Platforms

Resource Toolbox

Information Technology