Objectives
By the end of this session, trainees will be able to:
- explain why trainees sit at a higher AI risk point than senior lawyers
- identify red flags governing trainee AI use
- use AI without undermining supervision, learning, or competence
- disclose and document AI use in a defensible way
- apply a practical checklist to real trainee work
Content
Welcome, framing & ground rules
- Why this session exists (risk, judgment, learning — not tools)
- What this session will and won’t do
- Housekeeping & interaction format
- Why AI matters for trainees
- What generative AI is (plain English)
- Why trainees sit at a high-risk point in AI adoption
- Opportunities vs threats:
- The five non-negotiables governing trainee AI use:
- Confidentiality
- Accuracy
- Supervision
- Audit trail
- Competence
- Snapshot of AI tools law firms use (high level)
- Quick stories/concerns
- Can anyone share when AI has let them down or concerned them?
- Red Flag #1: Confidentiality
"Never put identifying info in public AI - ever"
- Why confidentiality risk is amplified by AI
- What counts as “identifying” (often broader than trainees expect)
- Public vs enterprise tools
- Consequences: regulatory, contractual, reputational
- Exercise
- Classify anonymised prompts as Red / Amber / Green
- Discuss why borderline cases matter
- Red Flag #2: Accuracy
"AI hallucinates case law, statutes, deadlines - always verify"
- Hallucinations: law, citations, dates, procedural steps
- Real-world examples
- When AI is useful vs dangerous
- Output review exercise
Triangulation of tools: review an AI written file for accuracy. Attendees can use AI to check it but must also non-AI tools to fact check (e.g. in-house research tools, Google search (not Gemini)
- Red Flag #3: Supervision
"Your supervisor must know when AI touched the work"
- Why undisclosed AI use breaks trust
- Supervision as protection, not punishment
- What good disclosure looks like (principles, not scripts)
- How to frame AI use when handing up work
- The “could I defend this later?” test
- 5. Red Flag #4: Audit Trail
"If you can't retrace your steps, it's useless"
- Why AI output without process is a risk
- What counts as an audit trail in practice
- Exercise
Build an audit trail for sample research task (prompt design task)
- Red Flag #5: Competence
"AI accelerates learning; it never replaces judgment"
- How over-reliance erodes judgment
- Difference between assistance and substitution
- Using AI to learn faster, not think less
- Reflection
"One competence rule for my future self”
- The Trainee AI Checklist
- One-page PDF
- Firm-brandable
- Explicitly framed as: “What we expect trainees to do if they use AI”
- Q&A, feedback & close
How To Book
Please contact your Learning & Development Manager if you are interested in this course.