Appendix E AI Disclosure and Use Requirements.docx
DOCX document 11 KB Posted
- Attached to
- CPRG Accelerator Research and Best Practices
- Solicitation number
- RFP EFAA 2027000006 5
- Issued by
- Denver County, Denver City, Colorado
About this file
This file is an Appendix to a state and local contract opportunity establishing Artificial Intelligence Disclosure and Use Requirements for personal service contracts and purchase orders involving research, analysis, and report writing deliverables. The document provides comprehensive guidance on responsible AI tool utilization while maintaining human oversight and transparency throughout the contracting process. The Awarded Respondent must disclose in writing prior to any deliverable submission whether and how artificial intelligence tools, including generative AI, machine learning, or automated analysis systems, were used in creating or contributing to the work product. Disclosure shall specify the particular AI tools or systems employed and describe the nature of their contribution, whether in research, drafting, analysis, or editing functions, and may be provided via memo, email, or disclaimer accompanying the deliverable.
The document establishes mandatory quality assurance and data protection standards requiring that all deliverables be reviewed, verified, and approved by qualified personnel with relevant subject matter expertise prior to submission. The Awarded Respondent bears full responsibility for the accuracy, quality, and completeness of all deliverables regardless of AI use and must ensure that AI-generated or AI-assisted content is fact-checked, reviewed for bias or errors, and validated against primary sources where applicable. Critically, the Awarded Respondent is prohibited from inputting confidential state data, personally identifiable information, or non-public information into any public or consumer-grade AI system, with any AI tools used requiring appropriate enterprise-level data protections. State data shall not be used to train, fine-tune, or improve any AI model without prior written approval from the State, and AI-generated content must be clearly distinguishable from original research and analysis with reports citing primary sources rather than relying solely on AI-generated summaries or conclusions for factual claims.
View the file
Other files
On GovTribe
Work with this file
- Download the original file
- Contacts named in this file
- Similar government files
- Ask GovTribe AI about this file
Text version
Appendix E: Artificial Intelligence (AI) Disclosure and Use Requirements Background and context The language below provides guidance on AI disclosure requirements and guardrails for personal service contracts and purchase orders (Pos) involving research, analysis, and report writing deliverables. The goal is to ensure the responsible use of AI as a tool while maintaining human oversight, quality control, and transparency.
AI Disclosure The Awarded Respondent shall disclose in writing prior to delivery of any deliverable, whether and how artificial intelligence tools, including generative AI, machine learning, or automated analysis systems, were used in creating or contributing to the deliverable. Disclosure shall include the specific AI tools or systems used and the nature of their contribution (e.g., research, drafting, analysis, editing). This may be provided in a memo, email with submission of the deliverable(s), or a disclaimer in the deliverable.
Human Oversight Requirement All deliverables shall be reviewed, verified, and approved by qualified personnel with relevant subject matter expertise prior to submission. The Awarded Respondent shall ensure that AI-generated or AI-assisted content is fact-checked for accuracy, reviewed for bias or errors, and validated against primary sources where applicable. The Awarded Respondent remains fully responsible for the quality, accuracy, and completeness of all deliverables regardless of AI use.
Data Protection and Confidentiality The Awarded Respondent shall not input confidential state data, personally identifiable information (PII), or non-public information into any public or consumer-grade AI system. Any AI tools used must have appropriate enterprise-level data protections. State data shall not be used to train, fine-tune, or improve any AI model without prior written approval from the State.
Quality and Accuracy Standards Deliverables must meet the acceptance criteria specified in the Statement of Work. AI-generated content must be clearly distinguishable from original research and analysis where appropriate. Reports shall cite primary sources and shall not rely solely on AI-generated summaries or conclusions for factual claims.
File details come from the government source that posted it. Updated .