Anyone who has participated in LegalTech the past few years has undoubtedly noticed a significant increase in the number of artificial intelligence and machine learning vendors present. This year was certainly no exception and as I walked around the show floor, I began to wonder how AI vendors might sum up where they think it is headed in the near future.
With that question in mind, I recently reached out to some of the leading artificial intelligence and machine learning companies focused in the E-Discovery and privacy law marketplaces including Ayfie, Canopy, Contexture and finally NEXLP for their insight. Around this same time, Forbes published the article “Will A.I. Put Lawyers Out of Business” which seemed to fit the subject matter perfectly as many in the legal field have begun to question A.I. in relation to their job security.
Artificial Intelligence Today – Excelling in data size and speed-to-knowledge
After speaking with industry experts, it was immediately apparent that artificial intelligence has not only already arrived but is already solving multiple problems in the E-Discovery environment. As Chip Delany, Strategy Associate for NexLP discussed, “We (humans) essentially have two problems; 1- productivity (time to completion) and 2 – bandwidth (processing of information).
NexLP’s ability to overcome these challenges today is through the creation of AI Models as a way of storing, understanding, and applying IP at scale. If you can leverage intelligence that is contextually accurate to your client’s needs, at great speed and scale, you have a distinct tactical advantage.” In other words, A.I. is currently lowering the amount of time and bandwidth required to sift through mountains of data and translate and the information at a higher level, in a hands-off and automated way. This equates to less human-input needed to perform remedial tasks like trying to reduce data noise or cull out non-relevant files while providing a much deeper understanding of the information within the dataset. In other words, relevant docs are identified much faster and with greater clarity.
Rob Wescott, Chief Strategy Officer at Ayfie added “the growth of data sets in E-Discovery continues to outpace budgets and human resources that can be applied to a matter.” This corresponds to Chip’s comment related to productivity and bandwidth in that artificial intelligence is being applied in order to identify and provide structure to vast amounts of data while “removing redundant information by identifying uniqueness and inclusive content.”
Ayfie is, “focusing on the identification of concepts in the text through context, leveraging natural language insights codified into the system through dictionaries, local grammars and extraction algorithms.” Essentially this technology allows AI to provide greater insight through language by providing quick access to meaning with the data. Context is analyzed to help extract and normalize entities layered in to help create the ability to extract entities (people, location, organizations, etc) which in-turn provides a more complete picture around custodian data. Translation – hands off, automated intelligence about data content.
Pete Fitzsimmons, CEO of Contexture AI said, “most at risk are those business areas that represent expensive and time-consuming challenges for organizations in managing and understanding content. Today, organizations of all sizes and types are all reeling from the exponential growth in unstructured content.” Unstructured data is certainly an area where both traditional processing of information as well as artificial intelligence and machine learning excel. The speed-to-knowledge and an understanding of data content removes a massive human hurdle in the need to sift and sort through mountains of information, run traditional keyword sets or queries and then having to review culled material.
Oran Sears, Chief Technology and Product Officer at Canopyco suggests that the incorporation of artificial intelligence has historically lagged behind other industries and needs to begin to quickly catch up. Studies have shown that unstructured data grows at approximately sixty percent per year. Being able to consume and make sense of unstructured data is key. Artificial intelligence allows for the creation of a “whole picture” in terms of entity extraction with layered levels of intelligence around key custodian identity and/or privacy information.
Heureka’s current part to play in artificial intelligence revolves around indexing, searching, and delivering data to AI/ML and analytics platforms. Heureka’s strength lies in our ability to target unstructured data sources giving users unparalleled access to information on desktops, laptops and file shares. Unstructured data is fast-growing and often acts as the biggest digital blind spot within an organization.
Intelligence gained using artificial intelligence or machine learning can be imported back into Heureka in the form of tags or classifications and applied at the file level using Heureka ACT™. By expanding the overall reach and depth of data, AI/ML engines can provide an even greater level of intelligence for any organization.
Artificial Intelligence Tomorrow – The road ahead
It’s difficult to make predictions in any technological field, and E-Discovery is no exception to that. One of the most rapid changes over the past several years has been the adoption of cloud processing, storage and review with companies such as Relativity, Zapproved, iCONECT-XERA, Logikcull, iPro and CloudNine embracing cloud processing and document review. Looking ahead, the benefits to storing data in the cloud will be significant as artificial intelligence and machine learning algorithms excel at analyzing large data sets. Cloud-based workflows can also easily take advantage of expandable high-speed, cloud-based processing engines (burstable performance).
Ayfie’s Rob Wescott sees AI as less threatening when thought of as “augmented intelligence”. “Those who embrace a future with augmented intelligence will be able to accomplish more with less. Some are ready to adopt new technology in their routine tasks, they will find more meaningful contributions.” He adds, “Not everyone will embrace these new tools. Those who do will work smarter, faster with more insight than others – augmenting ‘intelligence’ and their value to the organization.”
Automation saves time, increases efficiency and reduces cost
Automation saves time, increases efficiency and reduces cost. The future will be leveraged in many ways across the legal sector. For example, AI can play a role in creating standard templates, generating first drafts, identifying anomalies and alerting surfacing trends and patterns.
Contexture AI‘s Pete Fitzsimmons sees their Natural Language Processing engine performing greater tasks such as indexing, scoring and analyzing of text thus reducing the expensive and time consuming challenges that exist today for organizations in managing and understanding their data content. “In the next year, we expect to help our customers deliver new business models that will accelerate the time to value for content users, realizing the benefits of AI immediately, by enriching their content results and the user experiences continuously through machine learning.”
NexLP sees continuous improvements in AI models as the big near-future step. “The application [of AI] in eDiscovery and investigations has been for developing reactive solutions that use AI to find the smoking gun documents in a given dataset. However, if you are developing AI models that can learn from your investigations, that IP can be leveraged towards proactive monitoring solutions that sit on top of a company’s data and identify the risk before it become a downstream legal issue. The next step in Legal Technology is continuous improvement in AI Models that can be leveraged across legal operations to dramatically reduce cost whilst improving best practices and front-end decision making.”
Oran Sears from Canopy sums it up like this, “The list of areas where consumer uses of AI can translate into the Legal Industry is very long. That list shouldn’t scare the end users – AI will not put them out of a job. AI will make them more efficient. In my opinion, the Legal users should not only embrace AI, but demand it.”
AI will not put them out of a job. AI will make them more efficient
The subject of proactive monitoring is a place where Canopy strategy aligns. Oran Sears says, “If Netflix can suggest movies based on my watching habits, Google Home can be taught to understand my daily habits, and Amazon can tell me what products are normally purchased together, Legal AI should be able to help bring answers to questions the Reviewer didn’t know they had. For example, why not have a virtual assistant that can perform certain tasks on my behalf? Why not crowdsource decisions made by all reviewers to help suggest actions to be taken on review documents?”
The Heureka team is constantly evaluating the expansion of artificial intelligence and machine learning growth across multiple platforms including Google’s TensorFlow, Amazon’s SageMaker or Azure Machine Learning Studio along with automated detection systems like Amazon Macie, Google DLP, and Microsoft’s Compliance Manager/DLP. At the same time, we are partnering with companies who excel in the artificial intelligence and machine learning space and are leveraging their expertise to provide greater layers of intelligence in the Heureka platform via Heureka ACT.
Without doubt, unstructured data will continue to grow at a rapid pace as Heureka continues to expand our shared intelligence workflow model by providing seamless and rapid access to unstructured data, regardless of where it is stored. By doing so, we allow artificial intelligence and machine learning platforms to gain access to data stored in traditionally unreachable areas of a business.
Finally, Heureka’s expanded API’s allow intelligence platforms to interact with information in near real-time and provides the ability to supply tagging, classification and data intelligence at the file level regardless of where the file lives within the organization.
The future holds tremendous promise and opportunity as artificial intelligence and machine learning continues to expand across all markets. Heureka, along with our partners will continue to push the envelope to provide greater access, insight, and understanding in faster and more automated ways.