Instead of just providing a prompt, ask the AI tool what additional information it needs to improve its output. This "reverse prompting" forces the AI to ask clarifying questions, leading to more accurate and context-aware results in legal drafting.
AI performs best when given a focused task, such as revising a single purchase price adjustment section. It gets confused and produces lower-quality work when asked to mark up an entire complex agreement, highlighting the need for targeted application in M&A legal work.
Feed an AI tool all prior deal documents for a specific client to generate custom playbooks and internal benchmarks. The AI can chart how terms like survival periods or caps were handled across dozens of deals, providing powerful data for negotiations.
A client used AI to create a tax step chart for a reorganization, but the AI incorrectly assumed the target was an S-corp. This fundamental error, which could have cost millions, was caught only by a human lawyer's judgment, underscoring the necessity of expert oversight.
Some clients, concerned about data privacy, security, and accuracy, are explicitly opting out of having AI used on their legal matters. This requires firms to implement tracking systems to honor these requests and manage different workflows for different clients.
While AI can create a comprehensive list of issues from a contract, it lacks commercial judgment to prioritize them. It may red-flag a minor administrative point with the same severity as a major financial risk, requiring a human lawyer to filter the noise.
Beyond drafting documents, AI is highly effective at quality control tasks that humans often miss. Use it for proofreading, checking defined terms, and ensuring consistent formatting, which can catch subtle but important mistakes in complex agreements.
A valuable, time-saving application for AI in M&A is summarizing long and complex email threads. This allows deal professionals to quickly get up to speed on nuanced back-and-forths between clients and teams without having to read every single message.