Capitalizing on AI and ML as Your Legal Force Multipliers

Back to Blog Posts

Today legal practitioners are outnumbered and metaphorically outgunned when it comes to navigating the tsunami of potentially relevant data. Data volumes continue to grow at a breakneck pace with no signs of abating, yet the technology practitioners are using to manage and extract insights from data is often outdated and underpowered for the task at hand. What is a legal practitioner to do? 

There is a concept in military strategy that is especially apt in the current legal environment: the force multiplier. The Department of Defense (DoD) defines “force multiplier” as “[a] capability that, when added to and employed by a combat force, significantly increases the combat potential of that force and thus enhances the probability of successful mission accomplishment.” In the business context, the definition is even simpler: tools that help you amplify your effort to produce more output. 

Consider the difference between a military of cavalrymen with muskets and one of trained Navy Seals. The right tools amplify and empower a human to make more impactful actions and decisions. Applying AI and machine learning to legal technology reduces the cognitive burden to enable rapid, accurate decision-making. It also amplifies these decisions so fewer discrete decisions must be made. The result is greater data classification, with increased accuracy in a shorter period of time. 

AI and ML as a Force Multiplier 

This is not the first or the last time that you will hear the phrase force multiplier in the context of advanced machine learning and AI. In fact, the DoD started the Joint Artificial Intelligence Center (JAIC) as the DoD’s AI Center of Excellence to provide a critical mass of expertise to harness the game-changing power of AI. The DoD sees technology including machine learning (ML) and AI as a force multiplier as powerful as the musket or tank in prior centuries. The DoD and JAIC came out with the DoD AI strategy, which highlights the threat mitigation, force multiplication, and environmental factors driving advanced investigation and adoption of AI to protect the nation. 

One quote from the strategy is especially chilling and insightful:

AI is poised to transform every industry and is expected to impact every corner of the [Defense] Department, spanning operations, training, sustainment, force protection, recruiting, health care and many others... With the application of AI to defense, we have an opportunity to improve support for and protection of U.S. service members, safeguard our citizens, defend our allies and partners and improve the affordability and speed of our operations.

The DoD and many industries outside of legal see the impending effects of AI and ML as well as the strategic benefits of adopting broadly and early. In their mind, this is not something that can be ignored. No amount of burying our heads in the sand will change the adoption of technology and fundamental ways data is changing the world around us — even the historically glacial pace of federal government has had to catch up.

Applying Force Multipliers in Ediscovery

What does the military concept of force multiplier look like applied to legal practice? In the most basic terms, doing more with less. There are three major ways that technology is helping ediscovery professionals gain access to evidence with less human effort required, make better-informed decisions, and reduce time to insight. 

Early Case Assessment

Force Multipliers: continuous learning, automated batch refresh, advanced data visualization

Practitioners are seeing a force multiplier effect in early stage culling. Advanced early case assessment (ECA) dramatically reduces the number of man-hours necessary to get key evidence in even the largest data set. Additionally, tools like continuous learning models, automated batch refresh, and advanced data visualization enable a handful of practitioners to delve through a larger volumes of data — that previously took a basement full of reviewers — in a fraction of the time. 

Prior to the review phase, advanced data visualization enables case teams to identify gaps in their collection, prioritize key custodians and topics, and eliminate non-relevant information in a fraction of the time and expense of prior models. Rather than waiting for a human to touch each and every document, broad decisions are amplified across the entire data set to dramatically reduce the volume of data that even has to be reviewed. 

Data Investigation

Force Multipliers: continuous learning, data visualization, social network analysis, conceptual clustering, AI model sharing

Advanced data parsing and visualization that capitalize on AI and unsupervised learning models allow practitioners to gain insights into their data set in a fraction of the time. Social network analysis assists with custodian prioritization; data visualization and clustering assist with scoping and prioritization of a review; and continuous learning amplifies these insights across the entire data set.  

Additionally, new developments like AI model sharing allow practitioners to import lessons learned in prior matters or different data sets into their new matter. Rather than starting at zero, they have the benefit of prior insights and review hours to jump-start their analysis.  

Document Review

Force Multipliers: continuous learning, data visualization, social network analysis, conceptual clustering, AI model sharing

One of the most common examples of force multiplier at work in law is in the document review phase of ediscovery. Just a few short years ago it was not uncommon to have thousands of contract attorneys in a room going through millions of documents in a linear document review. 

Whether the team is using the insights from other matters by using AI model sharing or starting further along in developing their insights through social network analysis and concept clustering, the result in a drastic reduction of time to insight. Further, the AI-powered prioritization afforded by deploying continuous learning compounds this reduction of time to evidence.   

In general, a person would get through 50-60 documents an hour and there was limited ability to accelerate the timeline, aside from throwing more bodies at the problem. Today, teams leveraging advanced analytics and continuous learning see an increase in review rates of more than 60% compared to linear methods (that’s 88 docs/hour), and are able to more accurately get through volumes of documents.

The future 

Much in the way a tank amplifies human capacity in battle when compared to a person on foot, advanced analytic tools are broadly allowing leaner teams to reduce time to insight and dramatically reduce the cost associated with ediscovery. This trend is only on the rise. While not every case today takes advantage of even the most basic technological benefits (think email threading and AI-driven prioritization at a minimum), the day is not too far away when there will be a shift to the minority of cases not taking advantage of these — and even more advanced — tools. 

Subscribe to the blog
Cat Casey