Artificial Intelligence in Contract Management: Considerations for Practitioners (Part 1)

In case you haven't noticed, artificial intelligence (AI) is becoming a very big deal in business. And there is perhaps no area where the impact of AI systems (i.e., systems that exhibit intelligent behavior) will be felt more than in legal departments and, more broadly, in the area of managing contracts. Consider the enormity of collective knowledge that makes up commercial law and amount of money spent translating the semi-structured “legalese” that enshrouds what should be logical business constructs that sit at the core.

Every commercial contract is like a little knowledge base that contains critical data on organizational commitments (usually legal obligations), rights, remedies and rules that reflect business decisions made in the past that will affect performance in the future. Unfortunately, the amassed collection of thousands of these artifacts does not provide a “collective intelligence” that can be used efficiently to reduce commercial risks and increase economic value for the firm.

The trick therefore becomes not just how to digitize legal contracts, but also how to transform these documents into a structured commercial knowledge base that works in concert with AI based techniques and tools (of which the frequently mentioned term machine learning is only one of many) to create this collective intelligence.

But how? Does this require next-generation AI tools?

The answer is no (at least for now). But, if you want to build such commercial intelligence, whether on the buy-side for supplier contracts or more broadly for all enterprise contracts, you should take three basic steps:

  1. Build a high-level knowledge base about all your contracts in the form of a contract repository to gain high-level self awareness of your commercial health vis-à-vis your contract documents.
  2. Derive key intelligence from within your contract data to identify critical risks and latent opportunities (e.g., unclaimed money due to you). This is where the heavy lifting of AI starts by training the “machine” to decipher the legalese down to a granular contract clause level (including metadata).
  3. Begin to use your augmented commercial intelligence within upstream processes (e.g., strategic planning, negotiations, risk management) to plan, predict and optimize your commercial decisions at massively improved efficiency and effectiveness levels.

In this introductory post, I’ll explore the basics of AI and its relevance to Contract Lifecycle Management (CLM) so that practitioners can derive greater commercial intelligence from their contracts and set themselves up for the coming innovation in this exploding area.

Context: CLM 101

In a Spend Matters PRO Vendor Snapshot report on Icertis*, a CLM provider, I talked about the value of CLM, and I repeat it here to set the context and foundation for the subsequent discussion on AI:

“There is nothing more foundational, or business critical, than the contract. A contract is the ultimate commercial system of record, and as value chains become increasingly complicated and outsourced, the required business agility must get translated into commercial agility. Fundamentally, CLM systems help mitigate risk and prevent value leakage, but will also increasingly be critical to enabling new commercial relationships that support collaboration, new sources of value, and the total lowest costs to treat the risks within the relationship.

The [CLM] platform can be used to transform agreements from simply commercial artifacts containing legal obligations to “containers” of commitments that are used to satisfy stakeholders such as shareholders, regulators, and NGOs. The commitments can also be used as risk mitigations for various risk types that can also be modeled in the system. So, the CLM platform becomes a way to drive risk & compliance workflows alongside commercially related workflows that touch the contracts. All these requirements in turn dictate the need for an equally agile CLM platform that is functionally deep and highly configurable (e.g., APIs, analytics, workflow integration, data model extensibility, and so on).”

Whew, that’s a mouthful, but an important mouthful!

Now that I’ve set some context on CLM, subsequent posts in this series will cover the various areas of AI and explore how they are relevant and can be applied. Stay tuned!

*Note that Icertis is not an AI pure-play vendor within the contract management space, but rather a leading CLM suite with emerging AI capabilities. Exari is also a broad based CLM solution, and has proven AI-based contract analytics. The pure-play AI-enabled contract automation space is currently hotly contested by contract analytics vendors such as Seal Software (the clear leader based on the number of procurement implementations), Kira Systems, RAVN (iManage), Legal Sifter, LawGeex, Luminance, LegalRobot, Counselytics, and others.

Share on Procurious

Voices (2)

  1. Pierre Mitchell:

    ha! Jason, please remove your vendor hat and read more carefully. I never said iCertis was an AI based CLM solution (although they’re working on a lot of stuff you’re not aware of – as are a few other CLM solution providers – call me if you want to discuss generally – I can’t violate NDAs)

    I said.. “In a Spend Matters PRO Vendor Snapshot report on Icertis, a CLM provider, I talked about the value of CLM, and I repeat it here to set the context and foundation for the subsequent discussion on AI”.

    I was only pulling some text from a previous paper (that happened to be an iCertis snapshot) that was useful to set context and not have to re-type stuff. I’m very familiar with the whole LegalTech market/ecosystem, but this series is trying to distill the broad body of knowledge on AI and apply it to a specific domain which is buy-side contract management. I understand AI pretty well. I was coding in LISP in the mid eighties (when you were in grade school), developed a auto-classifier system, ran combinatorial optimization models, implemented object-oriented simulation tools, used multi-variate statistics (haven’t played around with SVM techniques yet though), and generally have stayed on top of developments in ML and AI.
    When you’re done reading the last part in the series and feel that I’ve mis-spoken about the topic, and would like to add to the body of knowledge that you feel corporate practitioners should tune into, I’d welcome your voice in the conversation.
    Thanks for writing in.

  2. Jason Gabbard:

    Icertis is not an AI solution, or a machine learning play. More thorough research of the field would be useful before generating content

Discuss this:

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.