Bots in Law

When I think of artificial intelligence in the law firm, I immediately hearken back to that epic Star Trek: The Next Generation episode, when the android named Data is put on trial to decide if he’s a person or a piece of property. (Spoiler alert, if you can have a spoiler for a show two decades old: in true enlightened-progressive Next Generation form, the judicial institution respectfully protects Data’s rights).

But in our own corner of space-time, the robot is more likely to be part of the lawyer’s team rather than the one on trial.

Though there aren’t charming androids walking about, artificial intelligence has sprung into every-day use for many of us. Students in ivy-league MBA programs are likely to find basic AI applications as part of their course load. And even if you’re not at Harvard, AI is powering your spam filter as we speak, so be grateful for the bots (or feel fear that the first conscious artificial intelligence could evolve from the mailroom drudge filtering out porn and drugs all day).

If the business and tech world are already flush with robots, what do attorneys need to know to keep up? One legal expert is already predicting AI will change our profession like email did, eventually becoming indispensable to each one of us and weeding out the dinosaurs who refuse to adapt.

And it’s already started.

A.I. comes into our firms any time a computing technology can learn, reason, communicate, and make decisions. It’s already used by attorneys in electronic discovery and due diligence. Other programs can sift through court decisions and filing data to give you predictions about the way judges and other lawyers function.

One promising area for legal A.I. is the birth of better search engines for case law, allowing researchers to ask questions in a natural language and receive back the cases that are truly most relevant. These technologies learn from use, and grow more accurate over time, and even tout their ability to instantly serve up the most similar case to the one you’re working on. Some A.I. search tech also promises to constantly monitor certain issues, so you receive updates with each new or recently-learned aspect.

In general, A.I. promises to be an uncomplaining paralegal or research assistant, who can function at lighting speed and never wastes his time on social media (unless your A.I. meets up with the Facebook A.I. — no telling what happens then).

But as we feel the wave of change coming down on us, some are wondering: is artificial intelligence coming to our jobs, or coming for our jobs? Should we be hopeful or afraid?

Bot Apocalypse!

Summon the four horsemen, because the end is nigh, some prophecy. Legal theorist Joe Patrice led with the headline: “BakerHostetler Hires A.I. Lawyer, Ushers in the Legal Apocalypse.”

Patrice later qualifies that his real worry is that using A.I. search aids is “disrupting the continuing education of lawyers.” Rather than spending countless hours figuring out how to search for the right terms, and how to scrape up the most relevant case law, bots are reaching the point where they can take our plain-English queries and serve up the precise stuff we’re after. “But all that mental drudgery is what transforms that young lawyer into the future top-billing legal eagle,” mourns Patrice.

I was thinking ‘Apocalypse’ meant something like the sun black as sackcloth and moon red as blood whilst a third of all living things died, but a challenge to our continuing education is also, I suppose, worth worrying over. Just don’t head to the bunker yet.

More dire fears are that robots will get to the point that they can replace an attorney completely. The massive accounting firm Deloitte recently released a report claiming that over the next 20 years in the UK, 114,000 jobs could shift from humans to computers. And just last fall, a legal scholar teamed up with a tech scholar at MIT to determine that if all lawyers started using available legal technology, it would cut their hours 13 percent. Lawyers who bill by the hour are obviously more inclined to wring their hands over this number than attorneys working as efficiently as humanly possible for a contingency fee. But even then: in the DIY Age, there could be reason to worry about job security when average people can use tech to meet their own legal needs.

There’s also fear about over-reliance on A.I. to do this work. Don’t forget that when bots were programed to design color names, they came up with Catbabel and Dorkwood; and when it was unleashed on recipes, A.I. thought “Swamp Peef” and “Cold Crockpot Water” were tasty dishes for we carbon-based life-forms. Leaning too heavily on A.I. without continuing to hone our own research skills could mean missing major case law that the computers haven’t deciphered yet.

But that’s been our lot with each development in search technology, and you don’t hear many arguing the virtues of good old-fashioned card catalogues. We tend to learn the limits of our technology, and shape our work accordingly.

Real Value from Artificial Intelligence

But if A.I. doesn’t call forth Armageddon, could it make your job easier? One big answer for me is in that figure up above, where legal tech can cut down your work by 13%. That’s great news for small firms working on contingency fees. As these technologies sharpen and spread, they could also level the playing field, making it so a scrappy lawyer without the backing of a plump piggy bank of billable hours could streamline their research and discovery processes and devote their time to building cases and interfacing with clients.

And there’s even hope in cases of regular folks turning to the internet to figure out their legal questions, rather than hiring a lawyer. The New York Times notes that people who do this are likely to be members of underserved populations, who normally would have no access to the legal system at all. Those who are able to get a lawyer are unlikely to hand their delicate legal questions over to Google instead anytime soon.

The secret to avoiding panic over A.I. seems to be the same as the secret to partnering successfully with it: respecting the robot’s limits. For instance: cutting edge A.I. legal research software called ROSS, touted as “the super intelligent attorney” and feared as a horseman of the Apocalypse, is neither. As attorney Steve Dykstra put it after using ROSS

[The program] doesn’t actually “answer” the question, but rather it gives you a much more refined starting spot. Thinking back to my early days in practice when I conducted legal research (poorly), I remember how often I spun my wheels. I would spend hours considering a question from one direction, only to realize it was the wrong direction. Then I would re-think my approach and spend another several hours digging through cases and texts before I came to the conclusion that I had taken another wrong turn. I would eventually stumble upon the correct approach and come to a reasonable answer. But not until I had, in many instances, wasted a bunch of my client’s money.

Seen in this light, this artificial intelligence is just a great search engine. Similarly, legal and tech scholars predict that while technology becomes better and more efficient at sifting through data for us, there’s no indication that they’ll be able to take over other aspects of our work, like “advising clients, writing legal briefs, negotiating and appearing in court.”

But even if there’s no cause for immediate alarm, it’s true that we remain fundamentally uncertain about what future technology will look like, and how it will affect each aspect of our lives, including our jobs. The closest thing to certainty that we have is the warning that ignoring new developments will hobble our ability to grow our practice. For now, legal technology looks more like the helpful Data than the sinister Terminator – but ignoring its growth in the legal arena could prematurely terminate your professional clout in the future.