How AI is changing the legal profession
AI has entered law faster than many in the profession expected. Large language models trained on legal corpora (Harvey AI, Casetext CoCounsel, Microsoft Copilot for Legal, Lexis+ AI) can now perform research, draft documents, review contracts, and summarise case law at speeds that challenge the economics of traditional legal work billing. The billable hour model, already under pressure from alternative fee arrangements and client resistance, is further disrupted when AI can produce a first draft contract or research memo in minutes rather than hours.
Document review in litigation (previously one of the largest sources of revenue for large law firms and a major employer of contract lawyers) has been substantially automated. E-discovery AI tools process millions of documents for relevance and privilege at costs that are a fraction of human review. This has already reduced headcount at some of the larger document review operations significantly.
Legal work that remains human
Advisory work, advocacy, and transaction management remain human activities that AI tools assist rather than replace. A senior M&A partner advising a board on a complex cross-border transaction is doing something that involves strategic judgment, relationship management, negotiation skill, and contextual understanding that AI cannot provide. A barrister making oral submissions to a judge, managing a witness in cross-examination, or reading the judicial temperature during a hearing is exercising professional skill that has no AI equivalent. A family law solicitor supporting a client through a divorce involving children and contested assets is providing emotional intelligence alongside legal advice in a way that requires a human practitioner.