In development
Specialist Advisors
Illustration using public documents; no client data; not a published workspace.
Built-draft review · trained software · FDA draft guidance · January 2025
Software review of a built draft
The workflow puts source-linked challenge questions to trained software modules. A person makes the decision; the modules support review with evidence-linked questions.
Considerations for the Use of Artificial Intelligence To Support Regulatory Decision-Making for Drug and Biological Products
For the proposed context of use of classifying participants as low risk for outpatient monitoring after dosing, does the credibility evidence show that the model reliably identifies those who require inpatient monitoring, with performance criteria and safeguards commensurate with its high influence and the potentially serious consequence of a false low-risk classification?
FDA hypothetical example · quoted“Which participants can be considered low risk and do not need inpatient monitoring after dosing?”
A person reviews the risk context and evidence and makes the decision.
Work list Awaiting a person
- Check the primary source against the intended use.Awaiting a person
- Record any source limitation before relying on it.Awaiting a person
- Ask a person to make the review decision.Awaiting a person
Source details and rights Open
- Considerations for the Use of Artificial Intelligence To Support Regulatory Decision-Making for Drug and Biological Products
Source date January 2025 Retrieved 16 October 2026
FDA says its website content is public domain unless otherwise noted. Check separately credited third-party material; no FDA endorsement is implied.
Rights / licence basis - FDA draft guidance PDF
Source date January 2025 Retrieved 16 October 2026
FDA says its website content is public domain unless otherwise noted. Check separately credited third-party material; no FDA endorsement is implied.
Rights / licence basis