Contract Management Blog

The Practicality of Artificial Intelligence / Machine Learning (AI/ML) in Contract Management

Of 650 Contract Professionals attending a Contract Management Conference, how many raised their hands when asked if they were currently using AI/ML?

The answer: 6 people in total raised their hands at an IACCM (now World CC) Global conference. If the same question was to be asked today, the numbers would unquestionably be higher; albeit, the “interest” in using AI/ML for enterprise contract management software has always been multiples of that practical usage number, often approaching 50% or higher. What does one make of these data points? Two years ago, the practicality of using AI/ML within a contracting process was not as achievable nor was it the priority that it is today. Furthermore, AI/ML technology was still in a nascent state, despite it being magnitudes easier, more intelligent, and easier to “train” since it first started to surface decades ago.

Why Does AI/ML Matter In Today’s Contract Management Process?

In short, the infusion of AI/ML functionality into the contracting process does what most other automation initiatives introduce at a macro level; efficiency improvements and risk avoidance/reduction. In the context of enterprise contract management software, this allows for 3rd party and executed contracts to more easily get loaded into the centralized repository with the accurate capturing of key data elements such as the counterparty name, contract expiration date and contract value. In this regard, efficiencies are introduced as otherwise, this would be a manual, time-consuming and error-prone endeavor. In terms of risk avoidance/reduction, the capturing and interpretation of risky language within a contract may be the catalyst for moving forward with or canceling a relationship or transaction that otherwise would be too risky.

These two factors are compounded further through M&A activity which requires that due diligence efforts mandate clarity and assessments of contractual commitments. In the absence of such risk scoring mechanisms achieved through AI/ML, organizations are left to laborious and error prone contract risk rating. This frequently is performed in a superficial capacity which exposes the organization(s) participating in such contractual relationships.

AI/ML Deconstructed For The Enterprise Contract Management Process

While the high level, value propositions of AI/ML in contracting is framed out above as efficiency improvements and risk mitigation, the practicality of it within the realm of contract management can be segmented as follows:

Ad-Hoc Contract Ingestion

As a Contract Manager or business user, I need to be able to easily drag a document(s) into the system and have it automatically create the contract record, upload the documents, associate them with the contract record and extract and store relevant metadata. I also need to easily know if something went wrong or if the process succeeded.

High-Volume Contract Migration

As a Contract Manager (and not an IT person), I need to take a large volume of contract documents and have the contract management software create counterparty records (or associate with existing counterparties), create contract records, load up the documents into the system, associate the document(s) with the contract record. As part of this process, key metadata will be extracted and stored into the system; which may include obligation data.

Contract Language Risk Analysis & Rating

When a document, typically, third party paper, is loaded into the CLM software, it should be able to use AI/ML to digest the document, interpret the clauses and compare them against standard clauses within a Standard Clauses library. An interface should be available to review the risk rating, suggested standard clause language and allow for redlined insertion of standard clauses in place of those that are provided in the document.

AI/ML Based Contract Search

As a Contract Manager, Risk Analyst or any other role that requires visibility into contracts (3rd party paper or your internal paper), there is an on-going need to have the CLM software allow for the searching and identification of contracts that meet certain search criteria. The value add with AI/ML is that the searching and identification of content is not necessarily dependent on the precise criteria defined in a contract query; rather, the AI/ML engine should be able to look for language that is similar and appropriately relevant to deliver more meaningful search results. For example, a new regulation may be introduced which requires that all contracts are identified and available in a report when arbitration is the preferred vehicle for disputes as compared to litigation. In this scenario, standard criteria for such a query may include the term “arbitration”. With an AI/ML powered contract management software the AI/ML engine may extrapolate further to include such contracts where adjudication or mediation is referenced; however, this happens without having to issue such specific variances on the searching.

It’s Not Just Contracts Dummy

While this may not be as surprising as the 6 out of 650 people reference in the beginning of this post, it does raise an important point that so many AI/ML contracting enthusiasts miss. Specifically, contracting systems and of course, AI/ML enabled CLM software are employed to generate and manage many other documents that extend beyond the initial and primary agreement. These may consist of SoWs, Change Orders, sub-contracts, Amendments, etc. It is in this regard that the “It’s Not Just” references come into play. Many people forget about how important these supporting documents are within the context of their respective contracts. In this regard, AI/ML should be fully exploited for all contract documents and not just the primary agreements.

clm software ai powered contract management enterprise contract management software

The AI/ML Explosion And Why It Matters To Contract Life Cycle Management

If one was to perform a search on “Artificial Intelligence” in Amazon today, the website would return over 10,000 hits. In short, AI/ML is rightfully “hot” and deserves the attention it is getting today and well into the future. In the context of 'AI powered contract management, most vendors are now incorporating it into their product offering in one or more of the deconstructed methods defined above. The reasons for this mass deep dive of interest and practical usage can be attributed to the following trends:

AI/ML Goes Mainstream

While the context of this article is about contract management, it can be observed that the physical or virtual perusing of technology or business conferences will have a large emphasis placed on AI/ML. The reasons stem back to the macro view of efficiency and risk avoidance; goals desired by virtually every organization around the globe. Historically, AI/ML technology was a micro-industry with a few small, focused and deep players. However, the importance of AI/ML has made its way on to center stage as global enterprise software companies such as Microsoft, IBM, Salesforce, Google and other comparable organizations are investing billions of dollars into AI/ML capabilities, be it through organic developments or acquisitions. As a result, AI/ML offerings and interest are growing at a phenomenal pace. This also means that the leading AI/ML companies are now delivering high-value capabilities which scratch the practicality itch.

AI/ML Training Model Improvements

Recipes, Training, Cookbooks, you name it, these AI/ML vendors have it. Factually, these are terms that AI/ML vendors do use to represent their training models. A training model is the generalized term used to have the AI/ML product consume a sampling of documents, in this case, contracts. Once consumed, the AI/ML engine will then go about in understanding common patterns which allow for effective data extraction and language assessments to take place. In prior AI/ML implementations, such Training exercises were programmatic, necessitated massive sample volumes, and required extensive configuration and refinements. While some of this is still true, the Training models are simpler, often with a polished UX and less demanding in terms of sample volumes such that heightened accuracies are experienced.

Contract Volume Explosions

While this is a non-techie and simple observation, it is still an accurate one. Contract volumes, including supporting documents, continue to climb and it is virtually impossible to effectively extract/categorize metadata and risk rate contracts in a more productive manner. Consequently, organizations of all shapes and sizes are looking for AI/ML to fill that gap cost-effectively in a cost-effective manner.


The impact of COVID-19 on our global economy can be observed by everyone. In the context of contracting, organizations are becoming much more rigorous in understanding what type of exposure lies hidden within their contract repository. To this end, there is a renewed interest to understand an organization’s real risks by using AI/ML technology to help the contract management process and identify latent risks.

Where Does A Contract Professional Go With This Newfound AI/ML Knowledge?

If your organization has elected to buy CLM software work with your vendor to better understand their AI/ML offerings. Don’t expect perfection or close to it with AI/ML technologies; rather, go into it with eyes wide opened on the practicality of what can really be achieved. Gain clarity on the specifics around the engine in use and if it was created by the vendor or OEM’d. If OEM’d, don’t rely upon a “black-box” mindset. Make sure you understand the underlying AI/ML vendor and compare in one form or fashion, against other such products.

Speak with Sales

Our Contract Management Specialists have no less than 10 years' experience helping clients address their contract related challenges. We'd love to help you, too! Get Started