info@trustbridge-compliance.com
Home / Consulting / Assessments / AI-in-GxP Readiness Index
The TrustBridge AI-in-GxP Readiness Index

Find your gaps before an inspector does

Twenty questions, about six minutes, one per screen. You will see where you stand today, the tier you are starting from, and a candid read on what to fix yourself first, so your AI plans rest on ground an inspector would recognise. Your result appears on screen the moment you finish.

Prefer a plain single-page version, or did the interactive check not start above? Open it here.

Answer what you can, add your details at the end, and your scored read arrives by email.

1. How many AI or machine-learning use cases are live or in pilot in your GxP processes?
2. What is the highest level of autonomy AI has in your GxP decisions today?
3. How critical are the processes where you apply or plan to apply AI?
4. What is your 12-month ambition for AI in quality?
5. Do you maintain an inventory of AI used in GxP, with a documented intended use for each?
6. Is each AI use case tied to defined acceptance criteria before deployment?
7. Do you classify AI use cases by risk using autonomy and impact, not just GAMP category?
8. Is validation effort scaled to that risk class at the function level?
9. Are adaptive or generative models given extra controls, or kept out of critical GMP use, in line with the Annex 22 draft?
10. Is the integrity and lineage of training and input data verified and documented?
11. Do AI inputs and outputs meet ALCOA+ data integrity in the GxP record?
12. Do you validate the model itself (performance and boundaries), beyond conventional system validation?
13. Is there change control for model and version updates, with revalidation triggers?
14. Does every AI-influenced GxP decision produce a record of input, output, reviewer and disposition?
15. Is there meaningful human-in-the-loop review, proportionate to risk?
16. Could you explain and justify an AI output to an inspector who has never seen the model?
17. Do you monitor model performance and drift in production, with thresholds and actions?
18. Is there a defined ongoing state of control or periodic review for AI systems?
19. Is there a cross-functional AI governance body with clear accountability?
20. If inspected on your AI tomorrow, how ready are you?