Why legal AI tools go unused: Four common mistakes and how to avoid them
The market is flooded with generative AI tools promising to change the way your legal team works. But what happens once you have onboarded that shiny new product? A recent survey of large law firms in the US found that the average firm had 18 gen AI tools available for attorneys and another six in pilot. However, only about 20% of lawyers at those firms regularly use the available tools.
What is often missing is a clear implementation strategy that aligns with your firm’s workflows and leverages your underlying content. No one wants to onboard an expensive product that is rarely or improperly used. What are common mistakes that legal teams make when choosing a new generative AI tool?
Focusing on technology instead of workflows
Don’t start with the shiny new product, start with your work. Where are you losing time? Where are write-offs piling up? What tasks are repetitive, inconsistent, or unnecessarily manual? Once you have analyzed inefficiencies in your workflows, you’ll understand the root problem you are aiming to solve. Only then can you accurately evaluate whether a generative AI tool is the right solution for your team.
A litigation team might find that associates spend hours each week summarizing deposition transcripts in inconsistent formats, and that partners routinely write down that time. That’s a specific, measurable problem. A tool that produces first-draft summaries in a standard template can then be judged against it: does it save time, and do partners trust the output?
Implementing AI without an underlying knowledge management strategy
Generative AI only works with the knowledge it can access. If your best precedents are scattered across inboxes, saved under inconsistent names, or known only to a handful of experienced lawyers, a new tool will not magically fix the problem. First, you need a KM strategy that addresses what knowledge matters in your day-to-day legal work, where it is stored, and how it is kept up to date.
If a firm wants an AI tool to draft first versions of NDAs, the tool needs access to the firm’s preferred templates and fallback clauses. If three partners each keep their own ‘best’ version on their desktops, the AI will either draw on outdated precedents or none at all. Agreeing on a single curated set of approved templates, with an owner responsible for keeping it current, has to come first.
Underestimating client, confidentiality and data risks
Legal teams cannot evaluate a generative AI tool on functionality alone. They also need to understand what happens to the information entered into it, where that information is stored, whether it may be used to train a model, who can access it and what contractual protections apply. Client requirements, confidentiality obligations, attorney-client privilege and applicable data protection rules should shape both tool selection and permitted use.
A client’s outside counsel guidelines may prohibit uploading its documents to any third-party AI tool without prior consent. If a lawyer pastes a draft agreement into a consumer chatbot to tighten the language, the firm may breach those guidelines, and potentially its confidentiality obligations, without anyone realizing until it’s too late.
Investing without a plan for change management
You’ve chosen the right tool for your firm, completed the risk review and security assessment, and clicked send on the official launch email. But weeks go by and still no one is using that new tool – why? Busy lawyers rarely change ingrained habits without a clear reason to and generative AI is no exception. You need a launch plan to ensure that attorneys at your firm understand how the tools fits into their workflows and how it makes their work more efficient. That requires a detailed training and communication plan, as well as clear sponsorship from law firm leadership – not just the click of a button.
Rather than announcing a new contract review tool firm-wide, one practice group might pilot it with a partner sponsor who uses it visibly on live matters, run short sessions built around the group’s own documents, and share concrete time savings after the first month. Lawyers are far more likely to adopt a tool when a respected colleague shows them it works on their kind of work.
Overall, choosing the right generative AI tool for your legal team is a critical investment that can meaningfully impact profitability. Get the most out of a new tool by understanding how it will fit into your workflows, organizing your underlying content, assessing risks up front, and rolling out with a clear plan for change management and adoption. With this foundation in place, you can turn your AI investment into a tool that attorneys use on a regular basis to provide quality legal advice that clients value.