Artificial intelligence (AI) is increasingly marching into the legal industry, and many attorneys are nervous about its potential impact. We should not fear the evolution of AI in the legal profession; we should welcome it. AI will never replace the value of creative, judgment-forming human attorneys.
As Yann LeCun, Facebook’s head of AI research, said, “Our intelligence is what makes us human, and AI is an extension of that quality.” Rather than envisioning AI as a victory of computers over humans, we should embrace it as a value-building tool that facilitates our work as attorneys and gives us space and time to focus on more meaningful and complex tasks. In that spirit, this article explains how artificial intelligence is transforming contract management through increased efficiencies, risk intelligence, and consistency.
What Exactly Is AI?
AI, sometimes referred to as cognitive computing, describes machine operations that mimic human brain processes by learning and making decisions based on sets of data. AI programs work in two ways. Computers either train themselves through trial and error, which is called unsupervised learning, or they learn from humans, which is called supervised learning.
In the first way, computer algorithms learn to perform tasks independently by analyzing data. Then they generate new algorithms from the results of their tests. Google’s AutoML project is an example. According to Google, Auto ML can build algorithms that identify objects in photos more accurately than services built by human experts.
In the second way, humans provide data and instructions to an AI program, which then scans the data to detect patterns and create classification rules. Here’s a simple example to understand how humans train AI programs. We could train an AI program to classify whether an image depicts a bagel or a croissant. First, we would provide the computer with a large number of sample images of bagels and croissants called a training set. Then we would clearly define each image as a croissant or a bagel. The AI would create a model from this information containing a series of algorithms about whether an image is a bagel or croissant. For instance, it may note the size of the object and whether it has a hole in the center. Once the AI completes its training, we would feed it new images of bagels and croissants to determine how accurately the AI’s model classifies the objects. If the AI’s accuracy is low, we could add data to the training set and adjust the algorithm (the set of rules it follows) to improve it.
How AI Works in Contract Management
In the same way we could train AI programs to distinguish between bagels and croissants, an attorney can train AI contract management software to classify contracts. To do this, an attorney provides a training set, such as a large number of purchaser and supplier contracts, to the AI.
The attorney identifies the contracts for the AI. Then the AI gathers information from the training set to help it classify future contracts. For example, it may analyze the language the company typically uses when forming a purchaser-friendly contract versus the language the company uses when forming a supplier-friendly contract. The attorney could also train the AI to require, reject, or request variations on specific language within various types of contracts.
AI is different from the contract management systems firms have used in the past to store and organize file sets. Cognitive computing, or AI software, learns how to complete certain tasks traditionally assigned to attorneys. For instance, it can discover patterns within a dataset (such as a group of contracts), examine and test data, and provide the outcome of those tests.
Increased Efficiencies in Contract Management
Traditional contract management leaves a lot of room for improvement. It’s laborious, time-intensive and often inefficient. Drafting contracts is a lengthy process. Even if you have a template, you must read each contract and edit it as needed to adapt it to a situation. Delays are common if more than one person is involved in the draft’s creation or approval.
It’s difficult for attorneys to get a general overview of all contracts and the information they contain using traditional contract management. Legal departments and law firms handle a large volume of contracts, and qualified labor is expensive. Thus, they struggle to extract necessary data into organized databases.
Additionally, in many legal departments, contracts are scattered among various email inboxes, cloud accounts and hard drives rather than organized in one database. Because of disorganization, contracting parties often miss deadlines, opportunities for cost increases and automatic renewals, which sometimes leads to costly errors.
Because contracts are the foundation for relationships between buyers and suppliers, and because they govern a large portion of the financial transactions of a company, inefficient contract management can lead to notable loss of capital. One study of Global 2000 companies found that poor contract management costs companies 9 percent of their annual revenues on average.
Lawyers can save a significant amount of time and money for clients by digitizing and compiling contracts into an AI program and setting criteria to classify and extract information. Machine learning (ML) is a subfield of AI where computers learn to make decisions and organize information using algorithms as they gain access to relevant data.
One way ML provides value in contract management is through search-and-find tasks. Similar to the example listed above in which an AI program learns to distinguish bagels from croissants, attorneys can train AI to identify types of documents. For example, you can provide an AI program with an extensive training set of non-disclosure agreements (NDAs) and teach it to look for specific language in NDAs. AI can learn to identify NDAs (and specific language in those NDAs) within a larger set of contracts. And from there, you can create rules for the data, such as “If this contract is an NDA, then send it to the general counsel to review,” or “If the NDA includes x, y and z provisions, then approve it for execution.”
Lawyers can also teach ML to recognize the type of agreement under review and provide suggestions for relevant provisions. For instance, if you frequently contract with a specific supplier, AI contracting software can recommend provisions you’ve used in past contracts with the same supplier. Or, if you form a deal with a party in a different country, the contract management AI may suggest you include provisions relating to cross-border transactions, such as those for choice of law, governing language and currency variances.
Greater Risk intelligence
To manage risk through contract management, an attorney must keep track of all the terms, milestones, and other important dates throughout the life of the contract. Lawyers can use AI in legal services to extract relevant information and notify attorneys of deadlines, thus freeing up time for more important endeavors.
For instance, in an AI program, you can create rules for deadline reminders, such as “Find all contracts with milestone dates, and notify me one month prior to each date.” As you may imagine, allowing a computer to keep track and notify you of relevant dates can save an incredible amount of time. AI frees attorneys to analyze and mitigate risks specific to each contract instead of trudging through boilerplate clauses in lengthy contracts.
Additionally, lawyers often need to foretell the future. They may need to answer the question: “If we pursue litigation against this supplier, how likely are we to win?” Experienced lawyers can usually answer this question more accurately than new attorneys because they draw upon a larger data set from their years of work. Yet AI can access and analyze a larger set of data than any human, and thus its predictions are often more accurate. For example, when humans trained an AI program on 200 years of records from the Supreme Court, the AI predicted future decisions more accurately than human experts. Consequently, lawyers may be able to save companies significant money by using AI to more accurately predict potential contract litigation outcomes.
Greater Accuracy and Consistency in Contract Review
Relying entirely on humans to draft and review contracts inevitably opens up room for error. When humans copy and paste from templates to create contract drafts, they may omit important sections or overlook problematic language. Lawyers working with contracts spend countless hours proofreading documents, ensuring all necessary clauses are present, and flagging anything that may pose a risk.
AI can help cut down on mistakes and costs caused by human error in reviewing. Some AI programs allow users to select specific clauses they will accept, reject or require. Thus, when an attorney uploads a new contract, the AI program will review it, determine which clauses are missing and flag problematic clauses for review.
Contract management AI also gives attorneys the tools they need to maintain consistency in the terms of their contracts. For example, if a company wants a specific definition used for force majeure across all contracts, AI software can ensure the same definition is used in every template and can flag instances in which the definition has been changed or omitted.
AI contracting software allows attorneys to extract key data from contracts to provide an overview of important terms. For example, for a company with a plethora of freight forwarder contracts, the legal department will want to keep track of each contract’s provisions on termination, cure periods, liability, and indemnity. Assigning legal staff to keep track of and organize these provisions is expensive and time-consuming. Yet AI contracting software can easily take note of and standardize these provisions, making it simpler for counsel to understand noncompliance situations and deal with them properly.
Over time, as law firms and legal departments become more familiar with AI’s capabilities, using this tech will become standard—akin to using autocorrect. Eventually, firms and legal departments will understand that AI frees lawyers’ schedules to evaluate risks and advise clients, leading to both improved quality of their services as well as cost savings that they can pass on to their clients.
Rather than replacing attorneys, AI will transform the role of the attorney. Lawyers will trade monotonous, repetitive tasks for more fulfilling work that requires the abilities that humans solely possess, such as strategic thinking, empathy, value judgments and creativity.