AI simplifies Change Reviews and Annex I maintenance

AI assists in change management in two ways: 1. Review support: It helps reviewers evaluate proposed PSMF updates before approval by checking whether the change in text and/or attachment appears factually sound, whether it still matches the source section data, and whether it remains aligned with relevant regulatory expectations like EMA GVP Module II. It also summarizes what is acceptable, what needs correction, and what needs follow-up, which reduces the time spent decoding edits manually. 2. Change Log support: It helps auto-generate text entries for Change Logs (Annex I) for every change in a matter of seconds. This is otherwise extremely time-intensive to capture manually, especially when the summary is prepared late during PSMF finalization. AI can gather data from within the system and produce a concise summary of the edits made.

No. AI is mainly a support layer within PSMF Manager, it is not an approval engine claiming to replace any executive roles. Final responsibility for deciding whether a suggested edit or comment is to be accepted, ignored or rejected remains with the human Users of the organization using PSMF Manager.

Yes. One of its core features is that AI use can be switched on or off anytime by the Admin of the organization. If it is switched on, it can be used for change reviews and change log (Annex I) suggestions and if it is switched off, the option will not appear for AI to be used during the change management workflow.

AI helps Users to check whether a proposed update still fits the structure, scope, and wording expectations of the relevant section. Prior knowledge of regulatory intelligence enables AI to check and suggest comments in case the changes are not aligned in coherence with regulatory expectations. That makes it especially useful for reviewing content that may sound correct but is not fully aligned with how the PSMF should be framed for that geographical region, based on local regulations to be followed.

It can flag factual inconsistencies, missing required context, scope drift, irrelevant additions, grammatical errors, or wording that no longer matches the underlying section logic. That makes AI-assisted PSMF review for changes extremely useful as a first-pass quality check. Where annex-linked logic is affected by the change, if a proposed revision creates a mismatch between the section wording and its supporting annex logic, AI helps surface that issue for User.

AI helps improve internal consistency before content is approved into the live PSMF. That means fewer weak edits, fewer avoidable mismatches, and a stronger final document when the system is tested under inspection. Additionally, the entries in Annex I (Change Log) have always historically been under intense scrutiny during the inspections, thus warranting an automated mechanism to ensure that all the edits made are accurately and contemporaneously logged with every approved change into Annex I.

Because the intended use of the AI feature in PSMF Manager is not open-ended drafting, it is focused to provide support inside a controlled PSMF workflow, with human decision-making retained at the end of each of the processes. This ensures the AI feature is aligned with regulatory expectations and GxP caution, as all AI inputs are logged and maintained as audit trails to denote subsequent human intervention when it comes to finalizing the changes or for decision making.