Sign up for our newsletter and get the latest big data news and analysis.

Artificial Intelligence and Natural Language in M&A: When Context Matters

In this special guest feature, Ned Gannon, President of DFIN’s eBrevia subsidiary, explores how AI is proving invaluable in the legal world of mergers and acquisitions, especially in contract analysis. eBrevia uses machine learning technology developed at Columbia University to automate contract review and data extraction for law firms, corporate legal departments, and auditing firms. Ned began his career as a corporate attorney and witnessed first-hand how costly and often inaccurate contract review was. He realized there had to be a better way to perform such tasks than to have bleary-eyed associates working until 2am. This was the impetus behind co-founding eBrevia, where he served as the CEO from its inception through its successful acquisition by Donnelley Financial Solutions (DFIN) in December 2018.

Law firms are at the cusp of fuller adoption of artificial intelligence (AI), to augment—not replace—the brainpower associates and paralegals bring to bear. Broad adoption of AI that can read, analyze and summarize thousands of pages in seconds will be a business driver in the legal industry.

AI is proving invaluable in the legal world of mergers and acquisitions, especially in contract analysis.  Depending on the size of the deal, there could be hundreds, or even thousands, of legally binding documents to review. Corporate charters, bylaws, and organization charts need to be examined to figure out how two entities can become one. Customer contracts, vendor contracts, employment agreements, leases, license agreements and other corporate contracts all must be reviewed to ensure there are no overlaps or clauses that would hinder a proposed merger. Historically, this review and summary process has been relegated to junior associates at law firms. For large deals, the amount of time required to review decades worth of legal documentation to find an anomaly or problematic contractual provision is daunting, and something AI solutions are much better suited for.

As AI becomes more adept at analyzing contracts, very significant levels of efficiency can be realized: the software can review thousands of pages in seconds to detect contract irregularities and extract information, reducing contract review time by 67 percent. And because AI is not limited by human reading and comprehension speeds, software is able to scan and analyze many more documents for risks than what would otherwise have been reviewed due to time or budget constraints. This also frees up paralegals and attorneys to reallocate those hours to higher-level work. Following the acquisition, by extracting information from current and legacy contracts, an AI solution can also ensure – with an extremely high level of accuracy — that key contractual data and commitments are not missed while helping the merged entity stay in compliance with its contracts and also identifying revenue opportunities and cost savings.

One of the key areas that AI is improving on constantly is a deft, nearly-human understanding of the nuances of language and identifying concepts rather than keywords. For example, in an M&A transaction, it is important to identify the concept of Change of Control in contracts as part of the due diligence process. This is a concept that could be expressed using language like “change of control,” “assignment by operation of law,” “sale of all or substantially all of a company’s assets,” or “merger.” Of course, in a merger agreement, the word “merger” might be used hundreds of times with only one instance applicable to the Change of Control concept. AI is able to take into account the context in which certain words are used and identify relevant legal concepts regardless of the vocabulary used to express them or where they might be buried in a contract while leaving behind inapplicable uses of an otherwise relevant word. Machine learning software is taught to identify a given concept by seeing examples of that concept annotated in a training set of agreements.

Learning concepts or turns of phrase is important not only to discover nuanced verbiage that could derail a deal, but also to be able to flag potential semantic sticking points in a contract. As attorneys, judges, and mediators know all too well, contract disputes sometimes come down to a single word that each party interprets as benefiting their argument. An AI solution that applies learned concepts and nuanced definitions that could change within a different scenario can analyze contracts and documents to flag these semantic gray areas for human review.

Quickly eliminating the haystack of thousands of pages of legalese to uncover one or two needles is only possible with AI. An AI solution can analyze 50 or more documents in under a minute, flag the most relevant portions that need attention, and send them back to the knowledge worker for further consideration. Because of the efficiencies and scale that can be realized in due diligence and contract analysis, AI is poised to become more commonplace in the near term.

There is a huge incentive to improve efficiency through technology: law firms of all sizes are under constant pressure to reduce fees, yet still give a level of service as if every client were a Fortune 500 organization. An AI solution would have a ripple effect on any law firm, freeing up attorneys at all levels to tackle the more complex aspects of M&A deals. For mid-size firms, or those small firms started by BigLaw veterans who want to do large deals, AI adoption allows them to level the playing field and compete with the bigger firms. Broadly speaking, AI tools will allow attorneys to spend more time practicing law and less time trudging through routine contract analysis.

Sign up for the free insideBIGDATA newsletter.

Leave a Comment

*

Resource Links: