T hough only 10 percent of law firms currently say they are using AI, many are starting to catch on that AI – or at least practical AI solutions – can transform the way they work for the better. By automating and optimizing many common tasks, practical AI enables lawyers to work more efficiently and deliver more value to their clients.
Here is a quick sample of tasks that were difficult to accomplish – or, in some cases, weren’t even feasible to tackle – before practical AI came along.
1. Classify Documents Properly
In a perfect world, a detailed profile is filled out for every new document that is created within a law firm. This document profile details what kind of document it is – for example, is it a share purchase agreement? An employment contract? – what industry it’s most relevant for, and scores of other useful attributes. Keywords and tags are dutifully entered to reflect the contents of the document, so that people can easily search on these criteria in the future with minimal effort.
Alas, we don’t live in a perfect world. More often than not, people at law firms are too busy to take the time to fill out document profiles, and so these fields remain empty. Without being properly tagged, these documents are then quite difficult to find in the future. The default assumption is that lawyers and other fee earners will spend time and effort curating their own data, but there are only so many hours in a day – and so it rarely happens.
Enter practical AI, which can help with automatically classifying documents without any human involvement. Working in the background, AI can examine an immense number of documents, identify what types of documents they are based on their content, and then classify them accordingly.
As a result, firms can better leverage all the knowledge and work product that they have accumulated throughout their decades of history without having to painstakingly go through and tag and classify documents on their own.
Making its vast trove of prior work product easier to search and re-use is particularly important for firms that are utilizing – by choice, or by client demand – fixed fee arrangements rather than the traditional billable hour. In a fixed fee arrangement, every bit of efficiency gain gets directly reflected on the bottom line, making the ability to leverage prior work that much more valuable and important. Before you can make good use of prior work, you need to actually find it – and that starts with proper classification.
2. Extract Relevant Information Quickly
Once information has been properly classified, which makes it easily searchable, you can start doing some very interesting things with it – like extracting the pieces of it that you want to reuse or otherwise leverage.
Practical AI enables law firms to automatically read, extract, and interpret critical business information from large volumes of documents and unstructured data. For example, law firm associates and other professionals who previously spent hundreds of hours manually extracting clauses or other key data points from documents – a process that is prone to human error – can now streamline these tasks with pre-built and self-trained extractors able to complete these tasks in less time with fewer errors.
An example is helpful here in painting a picture. Suppose a request comes flying across a lawyer’s desk to gather all the share purchase agreements that the firm has completed in the past two years where the value is greater than $10,000,000 and the opposing counsel was Firm X. Because all of the firm’s work product has been properly classified, the lawyer can easily find those documents and start pulling important pieces of information out of them. While it’s possible to extract clauses or other data points from documents that haven’t been classified, you greatly increase the accuracy and relevance of the information you’re pulling if you identify what type of document it is first. For example, there's no reason to try to extract the mortgage termination clause from a share purchase agreement because there's not going to be one. Practical AI makes it easy not just to extract information, but to extract it from the specific type of document that is most relevant to the task at hand.
Beyond using classification and extraction internally on their own work product, law firms can also use it to more efficiently perform an array of client-facing tasks.
Consider M&A due diligence. Rush projects like M&A require all hands on deck and the result can be overworked, unhappy lawyers and subpar work. With practical AI, law firms have the ability to intelligently review documents with greater accuracy and productivity.
During M&A due diligence, for example, a firm might need to review thousands of employment contracts to make sure there isn’t any hidden liability that could come back to haunt them. In the past, it was sometimes only practical for a firm to review a subset of these documents – say 30 percent, or some other representative slice – due to time and cost considerations.
With practical AI, firms have a tool that makes it feasible to review the entirety of the available dataset, without having to skip a large subsection of the documents – any one of which might contain an important outlier than contains a high amount of risk. Think here of the employee who, for whatever reason, has a clause in their employment contract that says they have intellectual property rights to the products a target company produces. That’s a nugget of information that any firm would hope to catch during the due diligence review.
Reducing overall risk and delivering better service to the client, while reducing the time and cost of the task, is a win for all parties – and it wasn’t possible before practical AI.
3. Identify Expertise More Accurately
Let’s say a firm just opened up a new matter involving real estate development in Abu Dhabi. Now they need to harness the expertise within the firm on this subject.
Today, in many firms, that process looks something like the following: the firm sends out a mass email to its entire staff asking “Who knows about real estate law in Abu Dhabi?” They might get a hit, or they might not. Maybe the person that actually is an expert doesn’t get the email, or gets it but doesn’t have time to read it, or simply doesn’t want to volunteer the fact that they have expertise in that area.
As an alternate approach, a firm could always run a keyword search on the biographies of all its employees, looking for keywords like “Abu Dhabi” or “real estate law.” The problem here is that employee biographies, much like the oft-neglected document profiles discussed earlier, are rarely filled out as completely or in as much detail as a firm might like.
Practical AI provides firms with a solution to the problem of expertise identification. It can find and analyze documents found on disparate systems across a law firm, and use authorship and other data embedded in these documents to determine who in the firm has expertise on a specific subject. In doing so, AI identifies subject matter experts based on implicit information and behavior.
For example, if people within your firm have used templates on property law written by one specific person hundreds of different times, you can be pretty sure that particular person is very good at that particular subject. Likewise, if someone has generated 75 percent of their billable hours over the past 2 years working on international property transactions, chances are they have considerable expertise in that area.
AI is able to surface that expertise simply by interpreting the various “signals” that are given off by lawyers every day as they go about their business. Identifying these experts based on their behavior – without any extra steps or effort required on their behalf – provides firms with quicker and better results than scanning biographies for keywords or “hoping for the best” with a mass email.
The Next Step
Practical AI provides a way to automate and improve document classification, information extraction, and expertise identification. It also allows firms to unlock the specialized knowledge and unique expertise that they have built up over decades, helping them differentiate in an increasingly competitive and commoditized market. More than giving these firms a leg up on the competition, using practical AI to handle these tasks lays the groundwork for analyzing the data that’s been harvested. Classifying the elements of past matters – what documents were involved, which experts were involved, and so on – makes it easier to create predictive models around how much a similar matter might cost, or what the outcome of a similar matter might be.
Think of these predictive capabilities as just one more task that practical AI will enable firms to easily accomplish moving forward. It’s an exciting next step that will benefit firms and clients alike. ILTA