Industry executives and experts share their predictions for 2022. Read them in this 14th annual VMblog.com series exclusive.
AI-Based Contract Intelligence will become an integral part of Contract Lifecycle Management
By Marie-Pierre Garnier, VP, Cortical.io
The value of Contract Lifecycle Management (CLM) systems is shifting from creating, storing and monitoring legal agreements to providing deeper insights into risks and opportunities.
Just surfacing basic information like name of parties involved or termination date, or locating Force Majeure clauses, is not enough anymore. CLM systems must be able to deliver advanced analytics and semantic search functionalities that give end users an in-depth knowledge of their contracts. The trend is towards augmenting existing systems with what is called Contract Intelligence – AI-powered solutions based on Natural Language Processing (NLP) that recognize and interpret non-standard language – e.g. clauses formulated with different expressions.
Traditionally, CLM solutions are built around six pillars: centralization (storage of all contracts in a central repository), analysis & reporting, authoring (contract creation), integration (connection with different apps and systems like CRM and BI), optimization (maximization of the business value from legal agreements). According to Forrester, defining the standard terms of a contract (parties involved, begin and end of engagement) represents only 10% of contract language. 90% of contract language is about responsibilities and liabilities – and this is where non-standard language is used: different people in different departments use different terms and expressions, different corporations have different templates, new regulations bring new vocabulary, etc. CLM systems often fail to deliver reliable results because they cannot deal with vocabulary mismatch (clauses with the same meaning but different words) or language ambiguity (words having a different signification depending on the context). That’s why contract review is still associated with a high proportion of manual labor, human beings, despite their propension for errors, delivering more accurate results that systems unable to understand meaning.
Next generation CLM software will be able to interpret and classify provisions based on meaning. CLM systems relying on keyword and rule-based search will be replaced by those understanding the semantics, that is those understanding the context of terms. Extracting key information like dates and amounts is a basic functionality of CLM systems. With Contract Intelligence, it becomes possible to understand descriptions of dates (“not later than ten business days after demand therefore”) or amounts (“equal to three percent (3%) of the shareholders’ equity of XY corporation”). Advanced NLP also makes it possible to recognize the similarity of phrases like “upon execution of this agreement” and “when the contract is signed”, although they use different wording.
The search capabilities of a CLM system are also crucial and too often criticized as under-performing by end users. The trend is towards user interfaces that enable users to search within a document or within the entire repository by combining both keyword search and natural language search. Another trending feature that gets a lot of interest from business users are inferences. Inferences are a means of efficiently categorizing extracted information, or of converting this information into company-specific jargon or nomenclature. For example, in an insurance policy, the user wants to quickly identify the funding method, in other words the party responsible for paying the premium: is it employer or employee funded? Is it a shared funding? With advanced NLP, CLM systems are able to translate a phrase like “the employer pays for the policy” as “employer funded”.
With the rise of Contract Intelligence, human intervention when reviewing contracts can be reduced to a minimum while, at the same time, the accuracy of results is improved. Intelligent CLM systems are able to redline different versions of a contract at the press of a button, flagging only those clauses that semantically differ and need action, and identifying relevant documents hidden in the company repository that users had forgotten or never heard of. Users of AI-augmented CLM solutions are amazed by the significant quality improvement of analytics results. On the other hand, automation drastically reduces the time spent to review a contract – from one hour to 15 minutes, as industry experts say.
Many CLM systems do not offer yet advanced analytics and semantic search capacities. In 2022, they will have to look into how to expand their systems to meet the new requirements of business users and support them in their digital transformation process.
##
ABOUT THE AUTHOR
Marie-Pierre Garnier, VP at Cortical.io
Marie-Pierre has been following the artificial intelligence and natural language processing markets for more than 10 years. She is particularly interested in the sustainability aspects of business software solutions and has written several articles about the subject. In her previous role at the Information Retrieval Facility, MP was facilitating the knowledge transfer between industry search experts and information retrieval scientists.
