Scenario Lab: Ethical Decision‑Making With AI
Scenarios help turn abstract rules into concrete habits. Work through these with your team or on your own.
Scenario 1: “Draft the client email”
Situation: A supervising attorney asks you to draft an email explaining next steps after a hearing. You consider using an AI tool to draft the email quickly.
Risks to watch: UPL, inaccurate statements, tone, and confidentiality.
Your task:
- Write a role‑safe prompt that uses only attorney‑approved outline points.
- List 5 things you will verify before sending the draft to the attorney.
- Decide whether this is low/moderate/high risk and why.
Scenario 2: “The hallucinated case citation”
Situation: AI generated a paragraph with a case citation that looks real. You cannot find it in your research database.
Your task:
- What do you do immediately?
- How do you communicate this to the supervising attorney?
- How do you prevent it next time?
Hint: Remove the citation, verify the legal point with real authority, and document the verification.
Scenario 3: “Sensitive data in the prompt”
Situation: You pasted an unredacted draft demand letter (with client identifiers) into a public AI chatbot before realizing it was not approved.
Your task:
- Follow the incident response flow (contain → notify → preserve logs).
- List who must be notified in your organization.
- List what information you need to collect for assessment.
{{UPLOAD_ASSET:incident_response_flowchart.png}}
Discussion prompts (optional)
- When should we disclose AI use to clients (if at all) under our policies?
- What tasks should be prohibited from AI use?
- What would “competence” look like for our team (training + approvals)?