The Legal Reality of Artificial Intelligence
Written by Michael McDonald, general manager of Veritone Legal
Today’s legal professionals are being buried beneath an avalanche of data. Every year, 350,000 new cases arrive in U.S. courts. Each of these cases generates massive amounts of information, with each lawyer managing an average of about 18,000 documents per year in 2015, up 50 percent from 2013.
Sometimes a single lawsuit can produce mind-boggling quantities of data. In just one case, FBI agents seized evidence amounting to 200 terabytes of information, equivalent to 17 billion Microsoft Word pages, or about 1.7 million trees, if those pages were printed out.
From law books to precedents, to emails, to phone calls, to evidence documentation, the quantity and variety of data used in legal proceedings keep multiplying. Often, concealed within this data deluge are invaluable nuggets of insight that could lead to a “Perry Mason moment” where the outcome of a case is turned on its head.
In many cases, it’s no longer practical from a time or cost perspective for a lawyer — or even an entire legal team — to sift through all the available information. Increasingly, the only effective way to navigate this sea of data is to utilize artificial intelligence engines that can ingest all kinds of content, from printed text to recorded phone calls, at speeds and scale that far exceed human capabilities.
This approach not only makes it feasible to process every bit of information in a major case, it also allows a legal group to focus on more critical tasks while producing a quality of work that would not have been possible previously.
Just as importantly, AI can scan content to monitor and help maintain compliance in various forms of communications. Using AI engines, firms can check employee interactions conducted via media including emails and recorded phone calls to check if their language and conduct complies with legal regulations.
By leveraging AI, legal groups can implement a proactive approach to communications compliance monitoring, constantly and thoroughly reviewing material in real time with a level of efficiency that would be impossible using traditional manual techniques.
Taking the Fast Lane to eDiscovery
Even when using e-discovery software to hasten the process of managing textual data, it still can take thousands of hours to weed through audio and video files. When lawyers perform these menial tasks, they are effectively reduced to the status of administrative assistants, making poor use of their hard-earned law degrees.
AI technology can transform every second or frame of audio or video content into a format that can be searched and analyzed for words, phrases, faces, sentiment and voice identification. It then can take that unstructured data and produce an index, or structure, within a matter of minutes.
This dramatically compresses the time and effort required for e-discovery and frees lawyers to do what they do best: litigating.
Terms of Compliance
Legal compliance departments are always monitoring employee communications for language that might breach the rules. However, this is a tediously manual and reactionary process. Compliance is also an area undergoing continuous change due to regulatory and policy updates.
Manually scanning emails for nefarious keywords is one way to head off potential compliance issues. However, this approach by itself doesn’t account for the massive amounts of text and voice files that also must be monitored. It also doesn’t incorporate the time-consuming process of ensuring each file is thoroughly reviewed, or the continued tracking of regulatory updates to ensure compliance with the most recent policies.
Because of this, even the biggest and most resource-rich compliance departments now are finding it impossible to constantly review material and stay on top of changes using human capital only. It’s no wonder then that more than one quarter of in-house lawyers surveyed said compliance is the biggest challenge their departments face.
That’s why compliance departments and legal firms are turning to AI to augment internal processes. AI can improve efficiency, weed out false positive results, cut costs and make better use of workers’ time and company resources.
Intelligence for Compliance
Huge volumes of data including conversations from phone recordings, chats and emails can now be analyzed using cognitive engines built specifically to understand noncompliant language. Using natural language processing, transcription, translation and voice and text analysis algorithms, cognitive engines can identify unusual employee behavior and flag terms so management can address issues. This contextual analysis helps save businesses money by avoiding fines.
Separating False-Positives from True Compliance Risks
Compliance breaches can paralyze legal firms, with organizations becoming bogged down as each incident balloons into a full-fledged case that must be thoroughly reviewed. Strict compliance guidelines spurred by tightening regulations have instilled great anxiety in firms, making them fearful of missing even the tiniest of risks.
Fear of missing even the smallest of risks has spurred an abundance of caution when it comes to auditing each case. This represents a major bottleneck that has been created by strict compliance guidelines spurred by regulations.
Organizations increasingly are resorting to using pattern-detection software to streamline the review process. However, this technique often results in many cases turning out to be false positives. These false positives waste time as team members are forced to resolve the case. They also can lead to other serious detection challenges during the review process. These challenges include data overload, resulting in missing the important instances; alert overload, i.e. too many false alarms; and gaps in the skills needed to quickly recognize a real event.
Focusing on Business-Critical Efforts Instead of Monitoring Communications
Organizations are opting to increase the amount of compliance-related positions to accommodate for growth and to adhere to the strict guidelines. In the banking industry alone, compliance and regulation efforts cost $270 billion annually and account for 10 percent of operating costs. However, as stated above, each case can take copious time to review before determining whether it’s a false positive. The time-consuming and mind-numbing exercise can make employees blind to differences. It also monopolizes their time, which could be better spent resolving true compliance events. Not only can AI assist in weeding out the false-positive cases, it can streamline the review process, allowing employees to spend time on critical business initiatives.
Keeping Up to Date
In addition to fending off potential compliance issues, legal firms and compliance departments are also obligated to understand and follow the most current regulations from around the world. This is a daunting and expensive endeavor. The cost of complying with these rules has risen 50 percent during the past three years as a number of new, country-specific regulations have come into play, according to the consulting firm Accenture.
Cognitive engines have been created that can understand and analyze the high volume of regulatory changes, and can ensure a business is aware of the most up-to-date policies. Natural language processing engines can parse regulatory text and pattern match with a cluster of keywords to identify the changes that are relevant to a specific organization in each geographic location.
As regulated industries strive to adhere to new rules and policies, AI will become an increasingly valuable tool for them. Cognitive engines can augment compliance-tasks, streamline processes and analyze data quickly and efficiently so employees can spend their time on real issues that impact a business’s bottom line.
Job Enhancement, Not Job Elimination
As a trade, the legal profession may be unfamiliar with the possibilities and improvements AI can provide when it comes to analyzing media examined in the discovery process.
The phrase “artificial intelligence” may inspire a Hollywood vision of an intelligent robot pleading a case in a courtroom. However, the reality is that a people-powered legal team can become smarter and more productive through the use of AI. Because of this, human lawyers will continue to play a central role in the practice of law. But with their AI partners on the job, these legal professionals will be able to focus their valuable time on crucial tasks that make the best use of their education and experience.
Michael McDonald is the general manager of Veritone Legal and has nearly 25 years of experience developing and implementing technology and software solutions for compliance, legal and related industries.