NDAs – The “Short” Documents That Matter
The Case for Artificial Intelligence and Enhanced Programming at Law Schools
Non-disclosure agreements (“NDAs”) are those “short” documents, typically one to six pages, that are used in many circumstances, including M&A deals, pre-employment interviewing, pre-collaboration engagement and litigation settlement discussions, among many other circumstances. NDAs are put in place to (1) ensure confidential treatment of proprietary and certain other information and (2) document certain other typical and customary restrictions on the actions of the parties to the NDAs.
Lawyers and clients alike know NDAs are necessary, but not many lawyers, clients or other professionals relish negotiating NDAs because of the sheer volume that have to be negotiated, particularly by corporate entities involved in large numbers of M&A transactions, collaborations or other commercial transactions.
So, how can we (1) negotiate, finalize and execute NDAs more efficiently, with less time and at a lower cost to clients; and (2) use NDAs as a tool to train junior lawyers and law school students, in each case using the latest available technology?
First, Question No. 1: How can we negotiate, finalize and execute NDAs more efficiently, with less time and at a lower cost to clients?
Consider NDAs in the context of an auction of a target company, business, or portfolio of assets (any of the foregoing, a “target”). Each M&A lawyer (and her client) who has ever worked on the sell-side of an auction is well-positioned to understand the need for NDAs to be handled more efficiently and how they can be used to enhance the training of junior lawyers and law school students. Each lawyer need only recall the countless days and nights she spent negotiating and finalizing NDAs, and the recollections will bring to mind how formulaic negotiating and finalizing NDAs are.
In a robust auction run by an investment bank, 50 to 100 or more NDAs could be negotiated on behalf of the seller. Typically, each NDA could take between 0.2 and 1.8 hours to negotiate, finalize, and execute, depending on (1) the form of NDA used and distributed to potential bidders or acquirors, (2) the extent of the mark-up provided by each potential bidder or acquiror, and (3) how many rounds of negotiations are required to finalize each NDA.
Assume that, for this set of circumstances, a third-year associate (working at a top-tier law firm located in New York City) is negotiating the NDAs at an average rate of $500 per hour (not taking into account the latest pay rises for associates). At that rate, and assuming the time frames above, each NDA could cost anywhere from $100 to $900 to negotiate, finalize and execute.
Now, assume that, on average, each NDA requires 1.0 hours of the associate’s time to negotiate, finalize and facilitate execution, or $500 per NDA. For negotiating 50 NDAs, the cost would be $25,000; for 100 NDAs, the cost would be $50,000. Then multiply those values, $25,000 or $50,000, by however many auctions a seller runs in a given year.
Query: How many auctions does a Fortune 500 company or a private equity firm run in a given year?
Answer: A lot!
The material issues that arise during negotiation of an NDA are generally the same from NDA to NDA. For example, the following provisions are often the subject of negotiation: (1) the definition of “confidential information”; (2) the purpose(s) and permitted use(s) of “confidential information”; (3) the definition of “representatives” and the responsibility for breach of the NDA; (4) the non-solicitation provision; (5) the “No poaching of customers or suppliers” provision (if applicable); (6) the standstill provision (if the target is public); (7) the “anti-clubbing” provision and certain other restrictive covenants relating to the availability of and access to financing; (8) the lengths of the terms of the confidentiality provisions and the restrictive covenants; (9) the provisions relating to disclosures required by law or requested in connection with investigations or similar proceedings; and (10) the return and/or destruction of confidential information (including the retention of confidential information for compliance purposes and carve-outs for automated back-ups or written record retention policies), among selected other provisions.
So common are the techniques and understood are the strategies used in marking-up and negotiating NDAs that a number of playbooks, “scripts” or “formulas” have been written to provide lawyers and clients with step-by-step guidance.
- For example, Igor Kirman of Wachtell Lipton Rosen & Katz wrote one such playbook in 2008 (https://www.amazon.com/Private-Equity-Confidentiality-Agreements-Line/dp/031498674X).
- Kirman, along with Nicole E. Clark of Vinson & Elkins and Saul H. Finkelstein of Ellenoff Grossman & Schole, put together a CLE webinar entitled “M&A Confidentiality Agreements: Strategies for Sellers and Buyers Negotiating Non‐Disclosure Provisions” (http://media.straffordpub.com/products/m-and-a-confidentiality-agreements-2011-05-18/presentation.pdf), sharing additional strategies for negotiating NDAs.
If playbooks, scripts, and formulas can be written and conveyed to lawyers and clients, then a computer program can be written to automate the playbook, script, or formula for marking up an NDA. And, if a process can be automated, a computer can be taught through defined computer algorithms, or machine learning, that automated process. And, there are many other examples of such playbooks, scripts, and formulas.
From Playbooks, Scripts and Formulas to … the Use of Artificial Intelligence
LawGeex (www.lawgeex.com) and representatives from the law schools at Duke University, Stanford University, and University of Southern California have, working in collaboration, demonstrated that AI can successfully and more reliably negotiate NDAs than their human counterparts! (https://mashable.com/2018/02/26/ai-beats-humans-at-contracts/#ikC6NPTTIkqo).
- “Comparing the Performance of Artificial Intelligence to Human Lawyers in the Review of Standard Business Contracts” (http://ai.lawgeex.com/rs/345-WGV-842/images/LawGeex%20eBook%20Al%20vs%20Lawyers%202018.pdf)
- https://www.lawgeex.com/AIvsLawyer/
Regarding the lawyer – AI showdown, LawGeex explained:
“Twenty US-trained lawyers, with decades of legal experience ranging from law firms to corporations, were asked to issue-spot legal issues in five standard NDAs. They competed against a LawGeex AI system that has been developed for three years and trained on tens of thousands of contracts.”
“Following extensive testing, the LawGeex Artificial Intelligence achieved an average 94% accuracy rate, ahead of the lawyers who achieved an average rate of 85%.”
Without a doubt, lawyers can leverage the computing power of AI to be more productive and to focus more time on being strategic and collaborative in the advice we give and the relationships with build with clients. We can enhance how we work and how we provide services and work products to our clients. Think about the cost savings alone that can be gained through implementing available technology to “deal with” NDAs.
Now, Question No. 2: How can we use NDAs as a tool to train junior lawyers and law school students, in each case using the latest available technology?
So, a couple of years ago, I drafted the attached outline for a clinical program, or lab, that is focused on NDAs for incorporation into law school programs. The impetus for drafting and sharing with law school professors was the hope that professors would help think through how this prototypical program could be implemented. While I knew that the process of negotiating, finalizing, and executing NDAs could be automated, I did not know for sure how machine learning tools or artificial intelligence (“AI”) could be used to in that process. Beginning in 2016, I asked a number of professors, computer scientists, and other professionals whether machine learning tools, or AI, was capable of the natural language processing required to negotiate and finalize NDAs. The reaction I obtained was mixed: Some were perplexed; some did not have time to be bothered; and some contributed their suggestions to the outline. (And, I am grateful to those who contributed).*
- The outline can be accessed by clicking on this link.
- *Contributors include, among others, Professor Michael Bloom of the University of Michigan Law School (https://www.law.umich.edu/FacultyBio/Pages/FacultyBio.aspx?FacID=bloomich and https://www.linkedin.com/in/bloomich/ ).
Given what LawGeex and other legal tech trailblazers have demonstrated, there is no doubt that we can develop new ways of sharing knowledge with junior associates and law school students. One place to start is to curate law school programming by introducing students to the legal underpinnings of NDAs, to the tools that the 21st century lawyer needs and to the engagement with, and deployment of, AI to enhance the practice of law.
So, my hope – my suggestion – in sharing this outline is that we work as a profession, together with our clients, to incorporate the practical use of tech tools such as machine learning and AI into law school programming. And, my question to our community is: How can we build a course that will evolve as AI evolves and that will incorporate the input of lawyers, clients, professors, and law school students alike? We, as a profession, owe it to ourselves and to the new entrants into the profession to stay abreast of available technology; we have an obligation to practice innovation.
About one year ago, I spoke with a current student at Columbia Law School, and she told me that her all-in budget (including tuition, housing and living expenses) for her three-year stint in school was estimated to be $270,000, which, without a doubt, is a material investment in one’s future. With that level of investment, students merit, and should receive, the best training possible; that is, the training that will provide them the ability to maximize the returns (financial and otherwise) on their investments.
If I were a law school student, and there was not a tech component (automation, coding, AI, etc.) in my curriculum, I would ask why, and I would do everything in my power to ensure that tech tools were incorporated into my educational and professional experiences. It’s your investment; it’s your education; and, you merit the best education your investment can buy.
#LegalTech - #PracticeInnovation - #FrontierOfTheLaw - #ArtificialIntelligence - #ReformLawSchoolEducation - #TransparencyInLegalEducation
