AI-scored risk register for security and compliance teams
Build a workflow with an intake page collecting a risk description, owner, and category. Send the risk to an LLM with a rubric to produce likelihood, impact, and an overall score with justification. Create or update a record in the risk register, deduplicating against existing risks. Validate inputs, handle ambiguous submissions by flagging for human review, and keep an audit trail of score changes. Output the scored risk entry and its register record ID.
What this prompt builds
Submit risks via a page, score them with AI, and maintain a living risk register in records.
The problem
Risk registers decay into stale spreadsheets because scoring is subjective and updating entries is manual. Inconsistent scoring makes prioritization and board reporting unreliable.
Solution and impact
This workflow takes risk submissions from a page or Workbench, uses AI to assign consistent likelihood/impact scores, and stores everything in a records-based risk register. Risk intake and scoring become standardized and the register stays current automatically.