The AI Learning Hub Journal

Read This First

What this course is — and what it is noteducational orientation for legal professionals — this is not legal adviceWHAT THIS COURSE DOES PROVIDETask suitabilitywhich work AI can touch, and how closely to watch itFailure modeshow these tools go wrong, and why it reads convincinglyA verification workflowthe steps that check output before anyone else sees itQuestions to ask a vendorwhat to establish before a tool touches client workPolicy building blocksthe parts a firm has to decide for itselfAll of it is orientation you can act onWHAT IT DOES NOT PROVIDELegal adviceNothing here is advice on your matters, yourclients, or your professional position.Jurisdiction-specific rulesDuties differ by jurisdiction and by regulator.The specifics are always local to you.A substitute for your obligationsYour regulator sets what you must do. This doesnot restate, replace or interpret any of it.Ethics duties appear here as common principles only —how they apply to you varies by jurisdictionTHE STANDING RULE FOR EVERYTHING THAT FOLLOWSYou remain personally responsible for your work product, whichever tool produced itOrientation, not advice — this sharpens the questions; your regulator and your judgement supply the answers
Fix the boundary first: this course builds judgement about the tools, it does not tell you what your professional duties are

This Course Is Educational Orientation, Not Legal Advice

Before anything else: this course is educational orientation about how AI systems behave and where they fit in legal work. It is not legal advice, it is not ethics advice, and it does not create any advisory relationship. Professional-responsibility obligations vary by jurisdiction and by bar — the duties described here are framed as common principles, not as the rules that govern you. Rule numbers, disclosure requirements, and permitted practices genuinely differ between jurisdictions and change over time. Most importantly: you remain personally responsible for every piece of work product you sign, file, or send, regardless of which tool produced the draft. No AI system, no vendor assurance, and no course — including this one — transfers that responsibility away from you.

  • Educational orientation only — nothing here is legal advice or ethics advice for your matter or your jurisdiction
  • Ethics duties are described as common principles; the binding rules, numbers, and interpretations are your bar's, not this course's
  • You own the output: signing, filing, or sending AI-assisted work makes it your work in every sense that matters
  • Verify anything consequential against your own jurisdiction's rules and your firm's policy before acting on it

Why the Framing Has to Be This Blunt

Most professional AI training is written as if the main risk is missing out. In legal work the demonstrated risk runs the other way. Practitioners have been sanctioned, publicly named, and referred to disciplinary bodies for filing documents containing citations that a model invented. Courts in multiple jurisdictions have responded with standing orders requiring disclosure or certification of AI use. This is not a hypothetical harm being raised for balance — it is the single best-documented consequence of generative AI in professional practice to date. That pattern is the core teaching material of this course, not a footnote at the end. If you take one thing away, take this: the tools are genuinely useful, and the failure mode is genuinely career-affecting.

  • The documented harm in this profession is real and ongoing — treat it as the starting point, not the caveat
  • Courts have responded with disclosure and certification requirements; those requirements vary and are still evolving
  • Usefulness and danger are not in tension here — the same fluency that drafts well also fabricates convincingly
  • Where this course describes incidents, it describes the pattern, not invented names, dates, or figures

How to Read the Rest of This Course

The four modules move from orientation to practice to the hardest problem to governance. Module 1 maps which tasks AI suits. Module 2 covers contracts and document review, where the technology is most mature and the precedent for algorithmic assistance is longest-established. Module 3 is the one that matters most: research, citation, and hallucination. Module 4 covers ethics, confidentiality, and building firm-level governance. Throughout, capability claims are stated conservatively and failure modes are stated plainly. Where a duty is described, assume the phrasing is a common principle and the specifics are yours to check. Where a tool category is described, note that no commercial product is endorsed here — vendor claims are for you to diligence.

  • Four modules: task fit, contracts and review, research and citation, ethics and governance
  • Module 3 is the highest-stakes material — do not skip it because research feels routine
  • No commercial legal-AI product is named as reliable or endorsed anywhere in this course
  • Treat every capability claim as something to test on your own matters before relying on it

Prefer slides, quizzes, and saved progress? Read this lesson in the library — free, no sign-up.