Ask a grounded question
Open the agent in its dedicated workspace or dashboard panel. Ask what finding needs review, what evidence is linked or how to prepare a response for its owner.
Inspect proposed actions
Review the source, affected object and wording before applying a structured suggestion. A person remains responsible for the edit and sign-off.
Record the actual decision
Use findings and evidence workflows to record the justification. A chat is assistance, not a verified evidence link or legal advice.
Example: asking what to do with a finding
A reviewer sees a possible gap in an AI policy and asks the agent, 'Which source supports this finding, and what should the owner check next?' The agent can explain the available context and propose a structured next action. The reviewer opens the cited material, edits or rejects the suggestion and records the actual decision in the finding workflow. If the needed document is absent, a confident-sounding answer is not evidence; the next step is to request the document.
Use the agent without losing control
- Ask about a named system, finding or requirement, not a blanket compliance verdict.
- Check the linked source and whether the agent's context is complete and current.
- Review wording, affected record and permissions before applying a proposed action.
- Document the human decision in the finding or evidence record, outside the chat alone.
Common questions
- Will the agent change a policy automatically?
- No. It can assist and propose structured changes; a person reviews and applies them.
- Can I rely on its legal interpretation?
- No. The agent helps organize questions from available context. Qualified legal and compliance professionals remain responsible for interpretation and sign-off.
Bring the question into a review
Explore a fictional Fencia walkthrough or tell us which system and evidence your team needs to review.