How AI is changing legal practice

AI has entered legal practice quickly — and it has produced one of the most vivid cautionary tales in the whole professional world: a wave of lawyers sanctioned for filing briefs full of citations that don't exist. Understanding both the genuine value and this signature risk frames the entire course.

A note up front: this course teaches responsible AI use in legal practice, including professional-responsibility rules — but it is education about ethics and practice, not legal advice. The rules are jurisdiction-specific and evolving. Use this to understand the issues and ask the right questions; consult your bar's rules and your own counsel for your specific situation.

The real value: AI genuinely helps with legal research, drafting, contract review, document summarization, e-discovery, deposition analysis, and more. Adoption has climbed fast — the American Bar Association's technology surveys show AI use among lawyers roughly tripling in a year (to around 30% by 2024), and Thomson Reuters found the vast majority of legal professionals expect generative AI to become central to their workflow within a few years. The efficiency gains are real, and they are pushing firms of every size to adopt.

The signature risk — fabricated citations. Generative AI produces fluent text that can include hallucinated case citations, quotations, and holdings that look completely real but do not exist. Lawyers who filed such material without checking it have been sanctioned — and the count is now well over a thousand documented cases worldwide (Module 3). This is the defining failure mode of legal AI, and it is entirely preventable.

Why the two go together: the same capability that makes AI useful — generating fluent, authoritative-sounding legal prose — is exactly what makes it dangerous when unchecked. AI does not "know" the law; it predicts plausible text. That means:

• Everything must be verified — citations, quotes, and holdings checked against primary sources, always.
• Competence is now a duty — lawyers must understand what these tools can and cannot do (this lesson's close, and Module 3's ABA Opinion 512).
• Confidentiality is at stake — putting client data into the wrong tool is an ethics risk (Module 3).

The honest framing for this course: AI offers genuine, substantial efficiency for legal work — but it is used in a profession bound by duties of competence, candor, confidentiality, and supervision, where an unverified fabrication in a filing can mean sanctions, professional discipline, and harm to a client. So this course is verification-first: we cover the practical uses (Module 2), but the heart of it is using AI competently and ethically — understanding the rules and cases (Module 3) and building the verification workflow and tool choices (Module 4) that make legal AI an asset rather than a liability. Used carelessly, AI in law gets lawyers sanctioned. Used competently, with verification, it's a powerful tool — and competently is the entire point.