Bruce Orcutt from ABBYY shares how artificial intelligence can extract meaning from contracts in the same way as humans do, the challenges businesses face in using AI and the benefits the technology can offer.
Contracts are the engine of a business; they contain critical business intelligence needed to run the enterprise, yet enterprises continue to struggle to do something that seems so simple: connect contracts to business value.
Research by the International Association for Contract and Commercial Management found that inefficient contract life cycle management processes may cost organizations as much as 9 percent of their annual turnover. While poor contract quality and negotiation cycle times contribute to cost inefficiencies, an increasingly burdensome area is obligations and compliance analysis due to the prevalence of more onerous legislative and regulatory mandates.
For example, the Financial Accounting Standards Board (FASB) ASC 606 Regulation requires all companies public and private to comply with new revenue recognition rules based on the actual transfer of goods sold and services consumed, and then recognize revenue proportionate to what was actually delivered and consumed. This change in revenue recognition rules requires companies to re-evaluate their performance obligations.
Then, there is the impact of the EU General Data Protection Regulation. It mandates all companies to meet rigorous requirements to safeguard privacy rights of EU residents, institute technical and organizational measures to protect the confidentiality, integrity and accessibility of personal information and implement appropriate breach response plans to mitigate privacy risks. Failure to do so may result in substantial fines and legal remedies that may also impact corporate reputation. Not surprisingly, a recent survey by Ernst & Young found that the world’s 500 largest corporations are forecasted to spend as much as $7.8 billion on GDPR-related compliance activities.
AI for Contract Analytics
Extracting intelligence from contracts is not an easy task. There are many instances of contracts that span NDAs, lease agreements, trade finance transactions, loan agreements and service contracts. Analysis of terms within such agreements, especially those connected to risk and revenue, demand the application of automated processes, as analysis of performance obligations is an inherently labor-intensive and often error-prone task.
AI for contract analytics includes machine-learning-based document understanding building blocks that extract meaning from documents much the same way as humans do:
- Recognition that intelligently extracts, classifies and serves critical data from incoming image, email and document streams. Once captured, business-critical data is automatically validated for accuracy and integrated into corporate information systems such as contract life cycle management (CLM) systems, ERP and ECM applications.
- Entity Extraction that automatically identifies and extracts entities such as the names, organizations, locations, dates, quantities and monetary value from contracts.
- Natural language processing (NLP) that helps organizations infer meaning from agreements in context by analyzing the co-occurrence of contract clauses and their relationships within and between documents.
- Text clustering and classification that categorizes documents based on their similarity and relationship.
It may be surprising that the availability of automated processes notwithstanding, a significant majority of organizations do not have visibility into where their contracts are located – be it on shared drives, email servers, content management repositories or even in file cabinets. Moreover, contracts often enter an organization through multiple channels – email with PDF attachments, fax and electronic files. Document capture technologies, such as full-text OCR digitize documents, extract relevant contract metadata and map them into transactional applications such as CLM systems and to enterprise content management repositories, delivering efficiency gains and empowering organizations to improve contract management.
Contract Clause Segmentation
Contracts are inherently “semi-structured” documents in that while they may not have the formal, machine-readable structure of a database, they are created to provide logical structure to the document, as well as context for the content. Typically, contracts contain standard clauses such as duration, termination, indemnity, liability, confidentiality, jurisdiction, remedies and applicable law and jurisdiction.
While AI technologies such as entity extraction and natural language processing are advanced algorithms, the precision with which meaning and context may be extracted depends on the granularity of the ontologies used. But there is not a one-size-fits-all ontology for all contract types.
An innovative approach is document sectioning, which provides context for NLP systems to take a fine-grained approach to targeting each section within contracts, using micro-ontologies for clause and entity identification. The benefits of this approach for semantic analysis are immediate: a system that uses this approach and targets sectioned business documents can accurately identify business-critical information from documents and deliver higher precision and throughput.
There are proven and measurable direct and indirect benefits associated with the application of contact analytics.
Contract analytics enables organizations to realize as much as 30 percent savings in contract administration costs and 20 percent savings in costs associated with terms and conditions compliance. Improved visibility to compliance obligations also helps companies improve their negotiating leverage in the enforcement of rights.
It has been shown that for every $100 million of profit, organizations may realize up to 15 percent margin improvement through the implementation of contract analytics solutions and best practices.
Many contracts include price, volume, discount terms and auto-renewals. By gaining better visibility to these contractual terms, finance departments may, for example, exercise better leverage with their suppliers to extend early payment discounts, thereby improving both returns on invested capital and supplier relations.
Finally, the application of contract analytics empowers organizations to anticipate business volatility and institute measures to strengthen their competitive posture while also mitigating compliance risks.