Automation is the key to gaining efficiency in many industries. But for some types of work, basic automation, often referred to as robotic process automation (RPA), won’t suffice due to the complexity of the tasks. These can include assessing and analyzing content, finding patterns in data, and making decisions based on those findings.
The legal industry is one vertical market that is ripe for increasing automation, particularly with respect to the review and analysis of legal contracts. While some information in a contract can be considered standard or boilerplate, most contracts contain very deal- or situation-specific terms, conditions, dates, and language.
A hastily written or reviewed contract can wind up costing a company a lot of money. Busy law firms and legal departments generally do not have the time or resources to review every contract that is executed to ensure that every clause in a templated contract applies to a specific deal or agreement. If these standard clauses are not removed, it can affect the enforcement of a contract. Similarly, mistakes within a contract that are not found or corrected can also cause problems if the contract needs to be enforced.
Even when humans are able to review contracts, they have limits with respect to their ability to process information accurately. They get tired, can be distracted, or simply may not spot irregularities or errors due to a lack of experience reviewing contracts. That’s why artificial intelligence (AI) and machine learning (ML) are starting to be applied by legal departments and law firms to aid in this process.
Technology to the Rescue
There are several technologies that can be utilized to aid in the review and analysis of contracts. An older but still useful technology is optical character recognition (OCR), which is used to recognize the letters and words on printed contracts. An OCR engine can convert these characters to text that can be fed into a natural language processing (NLP) engine. The NLP engine can then process the text and apply meaning to that text.
By pre-training an NLP engine to recognize legal terms and phrases, a contract review application can quickly ascertain whether a contract contains incorrect or non-standard language. By incorporating subject-matter experts during this training process, the application can also extract and classify relevant information that can be used to analyze large batches of contracts with much greater speed and accuracy than humans. Further, anomalies can be automatically flagged and either sent for review by a human or corrected automatically if the contract-processing platform incorporates additional logic designed to process and apply corrections.
Other technology, including computer vision (CV), can be implemented to ensure that contracts to be executed have been signed in the appropriate places and that signatures match those on file (to prevent fraud). Deep learning (DL) can also be used to mine through troves of contracts to identify connections between contracts, as well as correlate the contracts with outcomes. This process can then pinpoint anomalies or issues with a specific contract or language within a contract that led to a positive or negative outcome, thereby providing guidance for writing future contracts.
Key Vendors and Case Studies: Saving Time and Money
Some of the key vendors in the space include eBrevia, Cortical.io, and Kira Systems, among many others. A common trait among most vendors is their marketing pitch: the use of ML will not only save time and money but will also allow legal departments or firms to review all of their contracts. Further, ML will allow these departments or firms to extract information and insights from these documents and then prevent costly errors due to mistakes, omissions, or other anomalies.
One case study highlighting how AI-driven contract analysis and review technology can be used comes from one of Cortical.io’s solutions, which was deployed by an audit and management consulting company to review lease agreements. By integrating Cortical.io’s Contract Intelligence solution into the firm’s contract-processing workflow, data extraction models were trained on just 50 sample lease agreements. Then, as the solution was deployed, relevant information was extracted from newly reviewed contracts, and this information was classified and incorporated into various financial reporting documentation. The result was an 80% reduction in manual review and data extraction time, thereby enabling significant cost reductions.
Another example of how ML is being utilized in contract analysis comes from Kira Systems. This company’s Quick Study platform can be used to reduce the time to identify and extract pertinent information within contracts, which can then populate a contract lifecycle management (CLM) or contract management system (CMS). The solution is pre-trained to identify common data points and clauses within sales contracts, vendor agreements, nondisclosure agreements, and other contracts. Because the system uses ML to identify metadata, the format and wording in contracts can vary and the system will still extract the correct information. This information can then be validated by humans to improve the model over time.
While these types of processes could be handled by humans, the incorporation of ML and automation means that a greater number of contracts can be read, analyzed, and reviewed with much greater consistency and accuracy. Scalability is also possible – without regard to workforce size or location.
Not surprisingly, vendors have realized that there is a huge opportunity for this type of software. Indeed, as Tractica notes in its report, Artificial Intelligence for Enterprise Applications, AI software used for contract analysis is expected to generate nearly $958 million in annual revenue by 2025, up from $85 million in 2019. Although there can be a ramp-up period required to train the models for a specific type of industry or contract, Tractica believes that the probable time and cost savings attached to AI contract analysis solutions are significant. These potential savings will generate substantial interest among corporate legal departments and law firms over the next several years.