Lexknights Pte Ltd (Reg. No. 201917412C) is registered and incorporated in Singapore. 

LexKnights is not a law firm and does not offer legal advice or services.

Terms of Use | Privacy Policy

© 2019 by LexKnights. All rights reserved.

  • Black LinkedIn Icon
  • Black Facebook Icon
  • Black Twitter Icon

The Potential of AI in Contract Management - Opportunities & Challenges

Updated: Nov 21, 2019



Artificial intelligence in emerging as a key enabler to making legal functions more efficient, effectively and value-creating. According to Zion Market Research, the global legal tech AI market will reach $37,858 Million by 2026 with a CAGR of around 35.94% between 2019 and 2026.

At present, investment in legal tech is on an all-time high. Legal technology solutions leverage AI to facilitate increasing complex automation as well as contextual delivery of deep knowledge and insights at scale. Contract management is one of the main areas showing great potential to use AI to deliver unprecedented value.

Under end-to-end contract management there are different modules where AI can contribute meaningful automation and enable smart operations.

Smart Document Assembly


Document assembly is essentially the process of putting together the first draft of a contract. Contract assembly software typically involves (i) a responsive questionnaire or input form to collect the user’s input with respect to the contract (e.g. chatbot interface); (ii) a contract template with embedded code (e.g. variables for names and dates; logic to determine whether to include certain provisions); and (iii) an assembly engine that applies the embedded contract code in accordance with the user’s input to generate a tailored agreement. The vast majority of the solutions require extensive upfront efforts (and resulting cost to customers) to design the responsive input interface and code the contract template.

Artificial Intelligence has the potential to drastically lower this upfront effort, which often poses a barrier to adoption:


#1 Transform precedents/templates into intelligent contracts.


Machine learning algorithms can be used to identify the areas of a contract that change from one transaction to the next, and even map dependencies - understanding that certain changes go together (e.g. removing a defined term that is no longer used after a clause is dropped). Effectively, this means we can use machine-learning to create an automatable contract template out of a set of past agreements.

This represents a step-forward from even existing automated contract assembly solutions, avoiding upfront costs of $300 per hour to convert an in-house template into an intelligent contract generator.


#2 Create questionnaires/forms with branched logic to acquire relevant customer input.


The next step is to craft the questions to extract the necessary input from the user. Machine-learning technologies such as natural language processing and understanding can be applied to this use case. Having identified the areas that change from one contract to another and the relevant dependencies, AI-enabled apps can then generate the questionnaires to be presented to the user for customization, including skip logic (based on dependencies) to deliver an optimized user experience.

AI-powered negotiation & review platform


Once a first draft of the contract is delivered, the agreement will often undergo a series of rounds of review and negotiation between the counterparties. The traditional approach to contract negotiations is highly manual, inefficient, time consuming, starting with transaction management.

Today, professional dealmakers are using generic software tools that are not fit for purpose to manage negotiations, making the processes around them highly error prone:

  • Each party saves multiple contract versions in standard local computer folders with often messy naming conventions;

  • Comments are exchanged in several forms via email, some being included as notes in the email body, others embedded in the contract;

  • Maintaining spreadsheets to track contract status, dates, contacts, and other pieces of data;


These methods may be manageable when low contract volumes are involved, the values are low and contract terms are basic. However, as volume and complexity increase, this approach becomes insufficient, prone to oversight and risk.

Contract management software enables you to establish a centralized electronic repository, or database of record, for your contracts, documents, and other records — making them easy to manage, locate, and secure. AI-powered contract management tools can also leverage NLU to generate status updates on negotiations, increasing visibility for stakeholders (e.g. noting % completion, which clauses are still being negotiated versus which have been accepted, etc). This in turn can be used by transaction managers to prioritize and keep negotiations on track.


In addition, next generation contract management software will leverage AI to deliver the kind of insights you may expect from a traditional legal professional contextually and at scale, helping you review the agreement to flag key issues for further internal discussion, recommend counter proposals and give you ‘ammunition’ to persuade your counterparty to agree with your position.


#1 Flag clauses that are not-market do not match in-house form or are invalid or infrequent.


Artificial intelligence can help to quickly triage large volumes of contracts, flagging clauses that depart from your company’s standard form or policy, that are off-market or which may be invalid. Different levels of importance can also be assigned depending on the importance of the clause where the deviation is found. Approval workflows required to accept the proposed language can also be encoded into contract management solutions in a rule-based workflow system.


#2 Suggest fallback language with contextual advise.

AI can also retrieve alternate language either from the company’s own precedents or a contract clause bank relevant to the point of negotiation at hand, giving the user a set of choices to counter-propose against a flagged clause. This can not only help unlock an impasse in negotiations with new formulas, but also speed up negotiations and, where sourced from a precedent, reduce contract language variance and reduce approval turn around times.

When your contract terms are reviewed with certain arguments, you need ammunition in the form of statistics and data to counter the terms and negotiate your points. Being armed with the relevant points /facts/numbers will help you to check context. Here, AI does the work of a lawyer - It helps in preparing and assessing a contract, and fact-checking to see what works in the scenario.


#3 Provide user statistics on frequency of different clauses within client & market.


AI-assisted data analytics can be a powerful tool when it comes to negotiations, allowing you to to provide solid, quantitative back-up for you negotiating position. An AI-enable negotiation and review platform can look across company precedents an anonymized third party contract to provide facts and statistics at a clause level, even taking into account the context of the transaction. For example, it may empower you to argue against an unduly lengthy non-compete providing you with the percentage frequency of inclusion of such a non-compete term on employment contracts for similar positions in the same industry and contemporaneous time period.


Approval Workflow Automation


Large multinational enterprises often have complex and comprehensive standard operating procedures, compliance policies and authorization levels when it comes to contracting. It is not uncommon for a contract to require the approval of various departments before signing and, often, who needs to approve and when does not depend merely on the type of contract, but rather goes down to clause-level details. For example, a sales representative may be able to sign a bog standard sales agreement without any additional approvals. However, a deviation as to service fee structure may require approval from the pricing team; a deviation on termination events may require approval from legal and compliance, etc.


This often results in one of two undesirable outcomes: (i) the sales representative is unaware and signs without the required authorizations exposing the company to future losses; or (ii) the sales representative takes so much time to navigate the organization to obtain the required approvals that the relationship with the client is damaged and / or the deal is lost.


Today, workflow automation tools are capable of being configured to orchestrate the necessary approvals on document level based on rules provided by the user manually. The next generation of approval workflow automation goes deeper with clause level understanding, determining the required approval process based on the specific provision that is subject to a deviation, and the gravity of it, saving valuable time and preventing compliance breaches with potentially devastating consequences.


Take for example, a loan agreement negotiated by a corporate relationship manager at a bank. While he may have authority over the deal structure, the anti-money laundering contract has very specific terms in it and the legal team is always concerned with any small change that may be suggested/recommended. If your client suggests changes in this clause, the AI can recognize that there’s a deviation suggested in a critical clause. AI automatically flags it, be it any channel (email, dashboard) and informs the legal team.


This is made possible by cognitive document automation (CDA) such as natural language processing (NLP) and machine learning, capable of clustering, and classifying documents and clauses, employing natural language processing (NLP) to understand text-rich, structured documents like contracts and other legal documents. NLP can be used to quickly and automatically extract key data such as contract dates, amounts, parties or addresses―anything of interest.


Contract Lifecycle Management & Integrations


Contracts are typically entered into to govern an ongoing relationship between the parties (e.g. lease agreements, loans, recurring services). Contract management does not end once the legal agreement is signed and in place. Ongoing agreements may be subject to modification, extension and or renewals (e.g. co-working membership agreements are modified when new employees are onboarded), the may prompt for certain notices to be given or actions to be taken upon specific milestones or deadlines (e.g. investment agreements may require the investor to issue a notice requesting conversion of notes to equity), or require active steps to renew at the end of each term (e.g. digital marketing services contracts or tenancy agreements, which require notices to be exchanged to renew before the end of a term. For example, lease agreements, loan contracts or recurring service deals may well have multiple events through their term, such as payments, notices and renewals.


These are important events that happen during the life of a contract and missing these can reduce in significant liabilities or loss of opportunities. Despite this, many companies fail to maintain systematic discipline around contract lifecycle management. Contracts are often filed into oblivion once executed and filed in such a way as to make it very difficult to retrieve the necessary information. Traditionally, firms manage contracts manually through file cabinet storage and folder. This practice reduces the overall efficiency of an organization and leads to time and resource wastage in collating, organizing information and can also cause misplacing information or wrong updates in the contract. The time saved in automation of workflow approvals can be utilized in other aspects which can add increase the worth of the company in other ways.


Moreover, archaic methods are often used to keep track of contractual rights and obligations. As an anecdote, a real estate agent working with a leading real estate confused to keep track of upcoming tenant lease renewals by putting an alert in their calendar two years from now because their internal system did not properly track renewal dates or notice periods.

To make matters worse, when action need be taken in respect of a contract, processes tend to be widely ad hoc and inefficient. For example, a leading global co-working space troubles their legal team to draft a bespoke amendment to a membership agreement when a member needs to onboard a new employee.


Artificial intelligence can be used to extract the relevant actions and triggers from each agreement, creating the necessary reminders, and even automating the generation of contract addendums, re-statements and notices and send them to the right parties on the right date. Moreover, by leveraging API integrations with other enterprise platforms such as HCM, CRM or ERP, the data extracted from each contract by the AI enabled contract management tool and quickly update all relevant systems to maintain accurate records and trigger other workflows as required.


To illustrate with an example, take a one-year contractor agreement. The HR user would receive an alert from the system two months before expiry of the contract asking whether the user wants to take advantage of the option to renew for an additional one year. The user clicks yes and, with that single click, the AI-powered system uses the extracted terms of the contract to generate an addendum of extension, triggers an e-signature workflow and connects with the human capital management system to update payroll instructions for the next year.


Challenges with Traditional ECLM solutions


Conventional approach to contract lifecycle struggles with manual processes, legacy technology and insufficient proactive action in managing the contract lifecycle.

In the past, organizations took different approaches to contract management based on the fact that IT would drive the requirements and procurement of a new CLM (Contract Lifecycle Management) solution. This led to creation of several disparate and disjointed, siloed solutions IT taking the easy off-the-shelf approach, only to discover that the solution cannot be configured to fit the organization’s needs. This is the reason why the traditional 'cookie cut' CLM software is no longer the tool of choice and flawed with costs, risks and user disappointments.


Moreover, Enterprise Contract Lifecycle Management (ECLM) was often overly focused on compliance and risk management to protect from downside, but insufficient emphasis has been placed on the potential upside that the next generation contract. Management tools can deliver. As awareness over this potential increases, and as legal, risk and compliance functions transition from being cost-functions to be expected by management to boost top line growth, new, smarter and more tailored contract management tools are being sought to create value at a business-wide level.


Bottom-line


There are plenty of extraordinarily powerful use cases for use of artificial intelligence in contract management, which truly have the potential to unlock vast amounts of value. It is early days, and the technology is not yet easily deployable. It requires large training sets and human rule-setting, and still delivers less than perfect degrees of accuracy. But the legaltech community is moving fast and working hard on making these use cases a mainstream reality, and tech-savvy companies are already experiment with the technology available to ensure they stay abreast of the competition. Early understanding the potential of AI in contract management is key to its successful implementation and execution to ensure you stay a step ahead. Right now, the AI tools offer the maximum and highest value to companies with large volumes of contracts - cutting time spent in contract review and drafting.


With automated contract management solution providers like LexKnights, companies can make significant improvements in their contracting processes and legal operations as well.

12 views