School of AI · Certificate programme

AI for Lawyers

Practical Foundations

AI is already changing legal work. Research that took hours can be summarised in minutes, and first drafts produced almost instantly. Used carelessly, the same tools invent case law and put client confidentiality at risk. This course teaches you to use AI productively, accurately and ethically.

Who this course is for
  • Advocates, legal associates and in-house counsel
  • Law firm partners evaluating AI for their practice
  • Paralegals, legal researchers and LLB students of all semesters
  • No technical background is required
Why it matters now

In Ishfaq Ahmed v. Mushtaq Ahmed (PLD 2025 SC 582), the Supreme Court recognised AI as an aid to legal research and drafting while warning against fabricated citations and over-reliance. In April 2026, the National Judicial (Policy Making) Committee issued National Guidelines for the Use of AI in Judicial Institutions.

What you will learn
  • What AI and large language models really are, and where they fail
  • Effective prompts for legal research and drafting
  • Summarising judgments, statutes and long case files
  • Drafting letters, notices and first-draft contracts faster
  • Verifying every output and avoiding fabricated citations
  • Protecting client confidentiality within professional ethics
Curriculum

Six chapters, eighteen lessons

01AI Foundations for Legal ProfessionalsChapter 1 · 2 sessions (A) · 3 sessions (B)

1.1What AI and large language models actually are

How these systems generate text, why they sound confident even when wrong, and what “hallucination” means in practice. Plain language throughout, with no technical prerequisites.

1.2What AI can and cannot do in legal work

Tasks where AI performs well, such as summarising and first drafts, set against tasks where it fails, such as recalling case law from memory. Realistic expectations before any tool is used.

1.3A tour of the tools

General assistants such as ChatGPT, Claude, Gemini and Copilot, compared with legal-specific platforms: what each is good at, what each costs, and which are practical in a Pakistani practice.

02Prompting for Legal WorkChapter 2 · 3 sessions (A) · 4 sessions (B)

2.1The anatomy of a good legal prompt

Role, context, task, format and constraints. Participants rewrite weak prompts into precise ones and compare the results side by side.

2.2Giving AI the right context

Supplying facts, jurisdiction, applicable statute and the role the AI should adopt, and why a prompt that omits jurisdiction will usually produce foreign law.

2.3Refining outputs through iteration

Correcting, narrowing and re-prompting to improve a draft. Asking the AI to critique its own output and flag where it is uncertain.

03AI-Assisted Legal ResearchChapter 3 · 3 sessions (A) · 5 sessions (B)

3.1Summarising judgments and statutes

Extracting the ratio decidendi, preparing case briefs and comparing judgments. Participants brief a reported judgment with AI, then critique what it produced.

3.2Building a research memo with AI

Issue identification, statutory interpretation and precedent mapping, assembled into a structured research memo with AI assistance and human judgment.

3.3Verifying citations and spotting hallucinated case law

The verification rule: every citation is checked against PLD, SCMR, CLC, YLR or PCrLJ before use. Case study of Mata v. Avianca, where lawyers were sanctioned for submitting invented citations.

04Drafting and Document ReviewChapter 4 · 3 sessions (A) · 5 sessions (B)

4.1Drafting letters, legal notices and first drafts

Producing and then correcting first drafts of correspondence, legal notices and applications, with human review checklists applied to every draft.

4.2Contract review and clause comparison

Reviewing agreements for missing or unfavourable terms, comparing clause variants, and building a reusable clause library.

4.3Summarising long documents and case files

Working with lengthy judgments, case files and bundles, including Urdu and English material such as FIRs and court orders.

05Ethics, Confidentiality and RiskChapter 5 · 2 sessions (A) · 3 sessions (B)

5.1Client confidentiality

What must never be entered into a public AI tool, how these tools handle data, and safer alternatives for sensitive matters. PECA 2016 and data protection considerations.

5.2Professional responsibility and supervising AI output

Canons of professional conduct under the Pakistan Bar Council rules, the advocate's responsibility for every filed document, the Supreme Court's position in PLD 2025 SC 582, and the National Judicial (Policy Making) Committee's 2026 guidelines.

5.3Basics of a firm-level AI usage policy

Drafting a short, practical AI policy for a chambers or firm: permitted tools, permitted tasks, verification requirements and record keeping.

06Practical WorkshopChapter 6 · 3 sessions (A) · 4 sessions (B)

6.1Live case exercise

A full workflow carried out in the session, from initial research through verification to a finished draft, on a realistic case file.

6.2Building your personal legal prompt library

Each participant assembles and documents a set of tested prompts for the work they actually do, ready to use after the course.

6.3Final assessment and certificate

A practical assessment of the skills covered across the course, followed by the award of certificates.

Scheduling

Two ways to take the same course

Both options cover all six chapters and all eighteen lessons in full. Only the pacing changes.

Option AOption B
PatternOne day per weekTwo days per week
Session length45 minutes30 minutes
Sessions16 sessions24 sessions
Total duration16 teaching weeks12 teaching weeks
Best suited toA lighter weekly commitment across a full semesterShorter, more frequent sessions and earlier completion
Start here

Free 45-minute orientation session

No cost and no obligation. See AI at work on real legal tasks, and join a course afterwards only if you want to.

Already registered through your college? Check your official WhatsApp group for the day and time.