Computational Thinking and AI in Law.

The eight core sessions follow a cumulative arc in four phases of two. They begin by decoding confident claims and then opening the black box, so that the machine's failures can be predicted rather than merely listed (Sessions 1–2); move to computational thinking as the way to direct an AI system precisely (Sessions 3–4); turn to verification — tracing claims to origin, treating provenance as a legal question, and drilling the authority check (Sessions 5–6); and close by applying all of it to legal work and to the professional duties that make verification mandatory (Sessions 7–8).

1. 01 · Sessions 1–2

   Decode the claim and the machine

   Evaluate any confident claim — a person's, a statistic's, a chatbot's — then open the black box, so every failure mode that follows is one you can predict rather than memorise.

2. 02 · Sessions 3–4

   Direct the machine

   The four pillars of computational thinking, taught as prompt-craft: decompose, recognise patterns, abstract, and give the model a procedure.

3. 03 · Sessions 5–6

   Verify everything

   Trace a claim to its origin, treat provenance as the legal question it is, and drill the four-step authority check until it is automatic.

4. 04 · Sessions 7–8

   Apply it to law

   Long judgments, doctrine and comparison — then the professional duties that make verification mandatory rather than advisable.

1. 01

   Distinguish between a statement, a fact, data, evidence and proof — and apply that scrutiny to any claim, human or AI-generated.

2. 02

   Explain in plain terms what a language model does, and derive from that mechanism why hallucination, sycophancy, noise and inherited bias occur — including automation bias, the tendency of professionals to over-defer to a confident machine.

3. 03

   Communicate with AI systems through structured, precise prompts grounded in decomposition, abstraction, pattern recognition and algorithmic thinking.

4. 04

   Distinguish a generative tool from a retrieval-grounded one, choose between them on provenance rather than fluency, and run the four-step authority check — exists, says that, still good law — as a habit.

5. 05

   Verify AI output and AI-flooded content before relying on it, and state what would be required to authenticate an electronic record in an era of synthetic media.

6. 06

   Use AI responsibly for legal research and analysis — summarising long judgments, tracing doctrine, comparing authorities — while independently verifying every citation and claim it produces.

7. 07

   Apply the professional duties that AI use engages — competence, candour to the tribunal, confidentiality and privilege, supervision and disclosure — and produce a verification trail that would survive scrutiny.

1. 01

   Session 1 — How Not to Be Fooled (by Your Brain or Your Chatbot)

   The opening hour hands you the course's master decoding tool — the Ladder of Misinference: statement → fact → data → evidence → proof — and demonstrates, on the room itself, that the cognitive shortcuts which mislead human judgment are the shortcuts that mislead a large language model.

   • —System 1 vs. System 2 thinking; anchoring and priming
   • —The Ladder of Misinference as a tool to interrogate any confident claim
   • —Why identical fact patterns can produce wildly different human judgments (judicial sentencing variability as a live example)
   • —How the same anchoring effects that sway juries and judges also sway an AI model's estimates
   • —The Indian stakes: cost orders, a recalled tribunal order, and a Supreme Court warning about fake AI-generated judgments
2. 02

   Session 2 — Inside the Machine — How It Learns, and Why It Must Get Things Wrong

   Session 2 opens the black box, and it opens it early — because everything the rest of the course teaches is derivable from one mechanism.

   • —Training data, patterns and prediction — what a model actually does, and what it never does
   • —Why hallucination is structural: it knows the shape of a citation, not the contents of any database
   • —Why the same prompt gives different answers (sampling), and why it agrees with you (helpfulness training)
   • —“Garbage in, garbage out” and algorithmic bias — COMPAS, State v. Loomis, and the recruiting model that learned to prefer men
   • —Automation bias: the human half of the failure, and the one the profession keeps paying for
   • —Knowledge cutoffs, generative versus grounded tools, and what “reasoning” models do and do not change
3. 03

   Session 3 — Thinking Like a Computer I — Decomposition & Pattern Recognition

   Session 3 introduces the first two pillars of computational thinking as directly transferable prompt-craft.

   • —Decomposition: breaking a large, ambiguous task into sub-tasks a model can execute reliably
   • —Multi-step and staged prompting, with every stage as a checkpoint
   • —Pattern recognition applied to the tool: what a model is systematically good and bad at, and why
   • —Pattern recognition applied to the task: precedent-analogy as pattern matching
   • —Few-shot prompting and teaching a model by example
4. 04

   Session 4 — Thinking Like a Computer II — Abstraction & Algorithm Design

   Session 4 completes the four pillars.

   • —Abstraction: stripping a request to what actually matters
   • —The specification template — role, task, context, constraints, format
   • —Algorithm design: giving a model a precise, ordered procedure
   • —Chain-of-thought prompting, and why visible working is checkable working
   • —Iteration discipline: explore variants, then exploit the winner
5. 05

   Session 5 — Truth, Provenance and Synthetic Evidence

   Session 5 turns the decoding skills of Sessions 1 and 2 into a repeatable verification routine, and then puts that routine to work on the problem synthetic media has created for the profession.

   • —The SIFT method — Stop, Investigate the source, Find better coverage, Trace the claim to its origin
   • —Lateral reading as a verification habit, and the emotional spike as a signal to slow down
   • —Deepfakes and synthetic media: how they are made and why they persuade
   • —The confidence gap, and the liar's dividend that follows from it
   • —Authentication as doctrine: electronic records, the certificate regime, provenance and chain of custody
6. 06

   Session 6 — Grounding the Machine — Retrieval, Tools and the Authority Check

   Session 6 is the craft session that makes the rest of the course operational.

   • —Generative versus grounded: composing an authority versus fetching one
   • —What retrieval actually does, in plain terms — and what it fixes and does not fix
   • —The strongest move in legal AI: supply the document and ask questions about it
   • —The four-step authority check — exists, says that, still good law, and the step everyone skips
   • —Choosing tools by provenance; where AI genuinely accelerates legal work and where it introduces risk
7. 07

   Session 7 — AI and the Law I — Judgments, Doctrine and Comparative Analysis

   Session 7 is where the computational-thinking skills of Sessions 3 and 4 meet a real problem of legal research: reading long, multi-issue, multi-judge judgments efficiently and accurately.

   • —Decomposing a lengthy judgment into issues, holdings and obiter before prompting for a summary
   • —Extracting the ratio with paragraph numbers; using the model as an index, not a substitute
   • —Tracing a doctrine across a line of cases, and verifying every link
   • —Where AI summarisation flattens: dissents, concurrences, per incuriam findings, distinguishing facts
   • —Comparative analysis — judgments, statutes and clauses side by side, and what it cannot see
8. 08

   Session 8 — AI and the Law II — Duty, Confidentiality and the Verified Work Product

   The final core session closes the loop between everything the course has taught and the professional obligations that make it mandatory rather than advisable.

   • —The cases that drew the line, read as one pattern of a single skipped step
   • —Competence, including understanding the tool's failure modes
   • —Candour to the tribunal: you certify what you file
   • —Confidentiality, privilege and data protection: what a public model does with what you paste into it
   • —Supervision, disclosure, and the verification trail
   • —Where the regulation is heading — and why agentic AI raises the stakes

Rubrics across every component reward correctness of verification, soundness of prompt strategy, awareness of bias and sycophancy, ethical handling (confidentiality, candour, disclosure), and reflective insight. Each graded component carries an AI-use disclosure declaration — you are required to use AI, and required to disclose and verify it.

For the briefs, rubrics, and integrity notes behind each component, see the assessment detail .