Change doesn't always come easy for a 124-year-old insurancecompany. Nor does it come easy for the most traditionallyconservative and insulated department within that company: Thelegal department.

Liberty Mutual's legal department is nonetheless engaged in atechnology initiative to modernize its operations — an effort thatwill require leveraging data, updating in-house software and a decisive shift in workplaceculture.

This ongoing effort is aimed at making the department morecost-effective and efficient. It also is an opportunity tostreamline one of the company's core business offerings.

Integral to Business

At Liberty Mutual, the legal department needed financial andoperational support from company executives and leaders to take onits modernization initiative, which can be challenging to secure inany corporate business setting.

“Capital expenditure budgets are [usually] extremely tight incorporations, so it can be a whole lot harder to free up thesebudgets for noncore costs like a legal department,” said BradBlickstein, principal at the Blickstein Group, a consultancy thatworks with corporate law departments and law firms.

But Liberty Mutual executives, along with most others in theinsurance business, now recognize the urgency to innovate in orderto remain competitive.

The Pivotal Role of Data

Liberty Mutual creates and stores ever-growing volumes ofunorganized data, from settlement information and contracts tosensitive client records. Its legal team knows that taming thisdata through identifying, indexing and enabling access to it is thefoundation for every technology process and product they seek todeploy. But this essential first foundational step does not happenovernight.

“To be able to build up a capability to understand unstructureddata takes a long time; it takes a really long time,” said JeffreyMarple, innovation director, corporate legal at Liberty Mutual, whois tasked with turning the team's technology ideas into reality andinstilling a tech culture in the department.

The company's data scientists have made their way into the legaldepartment over the past 10 years, and they have since becomecritical to its operations.

“If you think about the insurance industry in general, we selltwo products,” Kiran Mallavarapu, SVP and manager of legalstrategic services at Liberty Mutual, said. “One is the indemnity,and the other one is [legal] defense.”

Having experts mine and translate data into measurable metrics,Mallavarapu explained, has enabled the department to “figure outvery simple questions, such as how many cases have we handled, howmany cases were in a particular week, how many were in particularjurisdiction [and] how long does each case take.”

This data is presented to legal staff primarily through datametric dashboards, which Mallavarapu called some of “the mostimportant tools” for the legal team. “I've seen a realtransformation in the legal department just having that informationvisually available.”

The data these dashboards offer is not just helpful for what itindicates, but also for what it leaves out. Since the datapresented shows what measurements and metrics are possible in thelegal space, it spurs legal department managers “to ask the nextbig operations questions” about what new information can be minedand leveraged in the future, Mallavarapu said.

Though the specific data desired will differdepending on each group, the overall legal department has a generalblueprint for what it would like to see in the near future. Incontrast to what Mallavarapu classifies as “descriptive analytics,”which looks at “what happened, [and] why did it happen,” the legaldepartment is now working “predictive and prescriptiveanalytics.”

Updated Metrics, Benchmarks

“Generally, we try to provide the right information at the righttime for people that are making decisions on litigated matters,”Robert Taylor, Liberty Mutual's vice president and senior corporatecounsel, said.

And there are simple, low-tech ways in which this can beaccomplished. “That may come in a simple alert to someone from thelegal department as to the status of a particular case,” or analert to inform them “that this case looks like other cases of asimilar type, and that you should pay attention to it for thefollowing reasons,” he said.

While such alerts are invaluable for those throughout theorganization, they are far from the only way Liberty Mutual's legaldepartment is enabling access to data to support betterdecision-making. One method being employed for increased sharingand collaboration both within legal and with outside departments isits knowledge management system.

The system includes document management repositories to help thelegal department ensure there are specific controls on sensitiveand confidential documents handled by their staff. Taylor notedthat document management repositories “create proper securityaround documents that the practice groups need, but also at thesame time allow them to share and reproduce that material as it'srequired to be shared and reproduced.”

Jeffrey Marple is Liberty Mutual's innovation director forcorporate legal. He calls the system “key to [the legal department]doing their job every day. They need a place to put things and findthings easily, making the knowledge management system crucial.”

The department is also looking to free up legal staff to focuson more high-level work by automating mundane tasks. This includestools for document assembly, which Taylor noted “are veryintuitive” and can be deployed “by subject matter experts in astreamlined fashion.” These tools, however, may only be the tip ofthe iceberg when it comes to future automation.

Not to Be Forgotten: AI

Despite the amount of work needed to implement AI, thetechnology is slowly catching on in a growing amount of legaldepartments across various industries. The legal department atCisco, for example, is designing pilot projects that use machinelearning to automate certain workflows and contracting tasks. WhileCisco did not divulge the specific applications it is considering,machine learning can be used to easily draft and execute standardcontracts like nondisclosure agreements. It can also be deployed incontract review to “read” contracts to determine which clausesdeviate from an established standard and what the types of clausesor terms are within each contract. Meanwhile, the legal team atYahoo has turned to machine learning to automate classifying andtagging invoices to ensure the integrity of e-billing data, and byextension, e-billing metrics.

While Liberty Mutual's legal department hasn'tyet identified any specific areas for applying machine learning,they are optimistic about what artificial intelligence may be ableto achieve in the long-term.

“I have this thing in my head: I call it the 'Legal Jarvis,'”Marple said. The Marvel Comics superhero “Iron Man had an AI calledJarvis and Jarvis did everything — Jarvis brought him lunch, butwould also do some pretty serious computational stuff and give himadvice on certain things. I think there is maybe, down the road,room for a legal AI as well. Where it lives, who it talks to, I'mnot sure, but we're figuring that out. To me, an AI gets rid of thefriction of work and allows our attorneys and legal professionalsto do the work they should be focusing on.”

Shifting the Corporate Culture

A future with “Legal Jarvis” however, will take more than just afew technological leaps. On par with the data it mines and tools itdeploys, Liberty Mutual's legal tech initiative significantlypivots on its success getting tech-resistant insurance andcorporate legal professionals to not only trust technology, but tobecome innovators themselves. It is a big part of the reason thelegal department launched a “legal ideation and transformationgroup,” a type of mini-incubator team that encourages in-houselegal to think like tech disruptors, and oftentimes helps turntheir ideas into reality.

“Within that group, it's not just about us coming up with ideas,or interesting vendors to work with,” Taylor said. “The ideationprocess requires that we discover, develop and test new ideas. Butthat's not enough, because a new idea is only valuable if you canimplement it.”