Artificial Intelligence to drive contract lifecycle management

In early 2018, McKinsey Global Institute estimated that nearly 23% of a lawyer’s job can be automated one way or the other. Automation will tranform multiple aspects of legal tasks. But many experts predict highly paid lawyers on the top of the pyramid will spend most of their time on value added work that demands higher level of cognitive skills. Non Lawyers, administrative staff, and technology will perform more of the routine legal services.

AI, Block Chain, Machine Learning, and Natural Language Processing are finally making inroads in the legal industry. They have ceased to be mere buzzwords and are quickly becoming mainstream in the legal industry. Till now humans had an edge over their AI counterparts. Understanding what AI contracting tools can and cannot do is key to a company’s successful implementation of a Smart Contracting solution.

What if Artificial Intelligence gave you an edge in Contract Negotiations?

  • AI can help us understand the flow of a phone or a text conversation – think Natural Language Processing and Speech analytics
  • When a seller is engaged with a customer, AI can pick up through intonation in the customer’s voice, and make suggestions, allowing  to counter the customer with an offer or discount.
  • Chatbots are good at negotiating. Yes! They can be trained in hard bargaining from results of past negotiations.
  • AI can comb through conversational data across lots of sales reps to understand where there are coaching and training opportunities, value prop improvement, and product improvement areas.
Negotiation with chat bot.

Predictive AI for Contract Lifecycle Management – design considerations

  • Are your existing contracts highly routinized and template driven?
  • Is the stakeholder internal or external?
  • What type of relationship do you have with your customers? Strategical or Transactional?
  • What are the compliance risk for your organization?
  • Are your T&C’s streamlined? They can take on different meanings in different agreements.
  • How are you managing contract exception today? Are they solely based on language deviation?
  • How many approval rules are there in place today? Are they clearly defined?
  • Are most of your contracts executed with the same counterparty?
  • Finally, the purpose of an AI driven Contract Lifecyle Management solution should never be to eliminate the attorney. Rather it should be used to increase a lawyer’s efficiency.

Natural Language Processing (NLP) looks quite promising

Natural language processing (NLP) which is a subset of AI that deals with interaction between human and computers using the natural language. NLP is quickly making inroads by freeing up attorney’s time from routine legal tasks. From a rich library of contract agreements and contents, NLP will have the ability to analyze and create patterns from these documents. This will allow contract teams to identify when a contract language has deviated from their defined standard.

Non-material changes to an existing contract or minor amendments can be auto approved, thus freeing time for the attorney and paralegal. NLP can detect unusual clauses in a contract to identify risk in contract lifecycle management. Although, material changes to a contract will require a high level approval.

NLP powered Contract Lifecycle Management (CLM) software

The degree of materiality can be determined by having NLP learn from past contracts with variation in the language that have been accepted or rejected.

Constantly feed rich contract agreements

Watch out for the following warning signs

Just like any new technology, there will be potential limitations that can make adoption of an AI driven CLM solution. Can the solution adapt to the law which is constantly changing with society’s progress? This is the biggest question that limits the adoption of the AI driven CLM software . AI software becomes more robust when it learns from data, identify patterns and make decisions with minimal human intervention.

Risks with AI Driven Contract Management

AI systems learn from experiences, much the same way humans do. They use data from the past to predict the future. Predictive intelligence or Machine Learning which is based on neural networks or regression can predict what will happen next, based on information about similar events in the past.

Unless the legal staff is constantly uploading new agreements reflecting changes in the law, the AI solution will continue to pick up stale agreement clauses that may be outdated leading to inadequate protection for clients.

Human vs AI cognitive power

Another potential pitfall is the lack of foresight or planning, which is very critical for a lawyer when drafting a contract for his or her client. Deep Learning which one of the branches of AI, are good at pattern recognition. Humans are capable of forming abstract models which often equals “legal thinking”. As long as the AI lacks the ability to construct and synthesize abstract situations, it will be of limited utility. In the end the goal of an AI CLM is not to eliminate the lawyer, but to increase a lawyer’s efficiency by reducing non-value added tasks.

Extracting meaning from agreement is hard

T&Cs can take on different meanings in different agreements. They are highly contextual. Did you know that 500 most used words in English have an average of 22 different meanings? You can imagine the level of difficulty it would take NLP to get it right when analyzing thousands and millions of agreements.

The process of reading and understanding English is quite complex. English doesn’t always follow logical and consistent rules. For instance, what does this news headline mean?

Scientists evaluate cancer risk of US drinking water

Are scientists scanning for cancer cells in the US drinking water? Or is the public water supply in the US causing cancer in humans? As you can see, parsing English with a computer program is going to be complicated.

AI driven CLM adoption roadmap

AI just like human intelligence is fed by volumes of data. Data about known events are used to train AI models and predict future outcomes. Clear expectations must be set before embarking on artificial intelligence to drive contract life cycle management. To begin with:

  1. Load rich agreement content and metadata.
  2. Start nimble – Focus on quite specific document types, such as NDAs, Routine Agreements, and Privacy Policies. You can increase the range of documents dealt with as you gain more customers and traction.
  3. Leverage NLP into the AI driven contract management solution to get the most bang for your buck.
  4. Automate, automate, and automate – Automation has clear advantages. Maximize chatbots to respond standard RFP questionnaires. Norton Rose Fullbright, an Australian law agency released in 2017 a chatbot for data security and privacy converns. The firm used IBM Watson to answer standard questions on GPDR queries.
  5. Construct a library of pre-defined agreement clauses.
  6. Provide legal protection to your clients.

Leave a Reply

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