| PURPOSE | SAMPLE CONSENT LANGUAGE |
|---|---|
| All Purpose General Use | This research study uses {insert AI tool name} to {insert purpose, e.g., analyze survey answers, transcribe interviews}. A human researcher will verify all AI‑generated findings to ensure accuracy. |
| AI Transcription | We will use {insert AI tool name} to transcribe audio recordings of our interview. The transcriptions will be reviewed and corrected by the research team to ensure accuracy. |
| AI Data Analysis | We will use automated artificial intelligence (AI) software to analyze your study data. This software helps identify patterns, but all final assessments are reviewed by our human research team. |
| Patient Monitoring/ Wearables | Data from your wearable device will be monitored by an AI tool to detect potential health changes. These alerts are reviewed by the research team to determine appropriate action. |
| AI Clinical Decision Making | Some tasks in this study may be supported by artificial intelligence (AI) tools that generate clinical decisions. Qualified human researchers will review AI outcomes. |
| Quality Improvement | AI tools may be used to help check data for completeness, consistency, or potential errors. These tools support the research process but do not replace human judgment. |
Sample Informed Consent Language for AI Use and Development
Sample Language Only
How to use this resource:
Consider how AI is involved in your study.
Is AI being used as a tool (e.g., transcription or data analysis), used to train or develop AI systems, or both? Only include language that accurately reflects your study activities.
Determine what sections and sample language apply.
Not all examples on this page will be relevant to every study. Think of the AI informed consent sample language as your menu options. Choose the language that aligns with your study design, data inputs, and intended use of AI.
Edit and tailor the language.
Modify wording as needed to ensure descriptions are accurate, clear, and specific to your research. Avoid using generic AI language that does not reflect how your participants will experience the AI being used or developed.
Ensure consistency across the consent document.
AI‑related language should align with other sections of the consent form, including study procedures, risks, data sharing, and confidentiality.
Key Information: What is the Study About...? AI Used as a Tool
Key Information: What is the Study About...? AI to Train, Validate, or Develop
| PURPOSE | SAMPLE CONSENT LANGUAGE |
|---|---|
| General Data Usage for AI Training | Your study data (including {specify, e.g., audio recordings, survey responses, images}) will be used to train and refine our artificial intelligence (AI) models. These models are being developed to improve {state purpose}. |
| General Data Usage for AI Development | We are conducting this research to develop new artificial intelligence (AI) systems designed to {specify analyze imaging/predict outcomes/analyze text}. This involves using your data to train computer models to recognize patterns for {state purpose}. |
| Train/Develop AI for Clinical Care Improvement | This research is being done to {specify develop/validate/test} an artificial intelligence (AI) system designed to {state function, e.g., assist with medical image interpretation, predict treatment response, etc.}. The purpose of this study is to evaluate how accurately and safely the AI system performs and to understand how it might improve clinical care in the future. |
What are the Possible Risks and Discomfort?
| RISK | SAMPLE CONSENT LANGUAGE |
|---|---|
| Bias: General Data Input | While AI tools are helpful, they can sometimes make mistakes, such as misinterpreting words or showing bias. We have a human researcher monitoring the process to minimize these risks. |
| Privacy: General Data Input | In studies involving AI, the main risk is to your privacy. Even when your name and contact information (or other identifiers) are removed, AI systems can detect complex patterns that, in rare cases, could make re-identification possible. Safeguards are in place to reduce these risks, but confidentiality cannot be guaranteed. |
| Privacy: Combining De-identified Data Sets | Combining multiple de‑identified data sets may increase the risk of reidentification because AI systems can detect complex patterns across data sources. Safeguards are used to reduce this risk, including limiting data access and monitoring how data are combined and used. |
| Multiple AI Systems | Multiple AI systems are used in this study, there is a small risk that errors could be amplified, or data could be shared between systems in ways not anticipated by researchers. Human oversight is maintained to minimize these risks. |
| Adaptive AI Model | The AI tool used in this study may continue to “learn” from data or be updated as part of ongoing development. This means the AI’s behavior or outputs could change over time in ways that are not fully predictable. Human researchers oversee the use of these tools and assess whether any changes could affect study findings or participant privacy. |
Who Will See the Information That You Give?
| DATA SECURITY | SAMPLE CONSENT LANGUAGE |
|---|---|
| Third-Party Software: Local Device | We will use {insert software name} to collect or transcribe your data. The software runs on our local hardware. Your data will not be shared with the third party. |
| Third-Party Software: Cloud Service | We will use third‑party software to collect or transcribe your data. Your data may be stored, retained, and used for training by {insert software name} according to the Terms and Use Agreement. |
| No Open AI: UK Protected Enterprise Environment | Your information will be accessed only by the research team and by authorized staff who support the study through the University of Kentucky’s secure research systems. If artificial intelligence (AI) tools are used to help analyze or process study data, those tools operate within the University’s approved secure data environment. Your information will not be entered into public or open AI systems. |
| Umbrella Protocol for Minimal Risk Record Review | The University of Kentucky Center for Clinical and Translational Science (CCTS) and UK Health act as custodians of health‑related research data. Any use of artificial intelligence (AI) involving identifiable information in this study will occur only within the University’s approved CCTS secure data environment. Your information will not be uploaded to public or unrestricted AI platforms. |
Additional Considerations
| TEMPLATE SECTION | SAMPLE CONSENT LANGUAGE |
|---|---|
| Option To Withdraw | You can choose to withdraw from the study. Your care/compensation will not be affected. However, once your data has been inputted into our AI system and used for training, it may not be possible to withdraw or remove it. |
| Benefits: Commercial Disclosure | The AI systems developed using study data may have future research or commercial applications. You will not receive financial benefits from these developments. |
| Closing: Questions About AI | You may ask the research team questions about how artificial intelligence is used in this study at any time. |