Deus Ex Machina, Lex Edition

I, for one, welcome our new Computer overlords.

— Ken Jennings, upon losing to Watson on Jeopardy

Unbeknownst to most, one of the start-ups in the Big Data tech arena of law is a company called Lex Machina.  For the moment, it’s limited to the IP niche of law, but what it seeks to do is rather astounding:

We mine litigation data, revealing insights never before available about judges, lawyers, parties, and patents, culled from millions of pages of IP litigation information.

We call these insights Legal Analytics, because analytics involves the discovery and communication of meaningful patterns in data.

In other words, they purport to be able to cull from the pages of court documents the efficacy of strategies, causes of action, lawyers, venues, judges, everything.  Forget for the moment that papers don’t tell all, and often don’t tell much, but that enough paper, with the right analysis, should provide a statistically accurate picture that replaces the “art” of lawyering with science.

Josh Blackman, who is by no means an enemy to the cause of big data, raised four ethical issues with the Lex Machina approach (though, to be clear, it had nothing to do with Lex Machina itself).

The first is well-discussed. How are these data analytic firms structured? Specifically, do they run afoul of Rule 5.4, which blocks ownership of law firms by non-lawyers. The answer to this issue will depend on whether the nature of the activities these analytic firms are perform are considered legal services. If no legal services are being provided, Rule 5.4 doesn’t apply. If there are legal services being provided, until the rules are changed, Rule 5.4 applies. It is only a matter of time before several players in the field who are coy about the manner in which they are structured are going to have to deal with this. I’ll leave that issue here for now.

Most notable here is that Josh raises the “coy” piece, as Alt Law, businesses owned by non-lawyers who are providing legal services (at the moment primarily to corporations, which everyone assumes can handle their own issues) are law firms when they want to be and not when others come sniffing around.  They are lucky that there is so little concern by prosecutors and disciplinary authorities in the Unauthorized Practice of Law, or they wouldn’t stand a chance. But hey, what’s a flagrant violation of law in the face of money to be made, lax oversight and the future of law cheerleaders?

Second, these data analytics firms, whether deemed law firms or not, will have to confront potential conflicts of interests.

Both sides of a dispute can go to the website, pay for some analytics, and get an answer, neither knowing that the other side did the same.  Since the computer doesn’t play favorites, it might not matter. On the other hand, the data from one could influence the data to the other. It could suck to be the first to inquire if you serve to inform the second.

The third issue concerns legal malpractice.

Plug in your info and Lex Machina spits out its data. You rely on it, crash and burn. You can’t sue Lex Machina for legal malpractice, because it’s not a lawyer, but a machine. And might I toss in an additional thought that so many people forget about these days: GIGO. Garbage in, garbage out. Don’t blame a machine because you asked a bad question.

The fourth issue, and the other elephant in the room, is Unauthorized Practice of Law (UPL).

At the moment, Lex Machina is primarily directed toward lawyers, so that whatever it produces is filtered through counsel before it reaches the client.  If their output sucks, it’s the lawyer’s duty to recognize it and address it.

But this is changing.  Lex Machina now offers a sample of what it can do with a demand letter, and this concern goes to the interests of the client rather than the lawyer.

With Demand Letter Analytics, Lex Machina provides this report to help you better understand your demand letter.  The first three sections provide information about the litigation activity of the company, law firm, and patents in your letter. With this information, you can better assess the threat of litigation. The last section provides information about other law firms, attorneys, and companies that have been involved in litigation with the company or patents. With their experience, these potential resources may help you better understand the threat you’re facing.

As anticipated, it didn’t take long to shift from providing big data support to lawyers to offering custom answers to individuals that can have a deleterious impact on their rights and future.

But we’re not IP lawyers? Who cares that these techno-freaks want to make IP lawyers and clients bow to the computer overlords?  Well, it’s just a hop, skip and jump from IP to criminal defense, where the strategies we appear to employ, at least on paper, will be subject to computer analytics. What are the chances a suppression motion will succeed? Should I cop a plea or go to trial? What’s my lawyer’s won/lost ratio?

Stupid questions? Not really.  Inadequate questions that fail to take into account a wealth of factors that papers could never reveal? You bet.  But the computer overlord won’t give a damn, because it’s a computer. Nor will Lex Machina because it’s a business, and if its business model means your clients go to prison, it’s of no moment as long as they paid.

Watson may have beaten Ken Jennings on Jeopardy, but law isn’t as easily programmed.


10 comments on “Deus Ex Machina, Lex Edition

  1. Adam Ziegler

    As a lawyer, I want every insight I can get – whether it’s the lawyer next door who has 30 years’ experience or the legal analytics software that sifts 30 years of data. I’m not naive about the limitations of either source of insight, man or machine, but that doesn’t mean there’s nothing valuable to be learned.

    As for “demand letter analytics” and the concern that this might veer toward legal services, it sounds like little more than historical facts and relevant context that will help clients identify a suitable lawyer and then arm that lawyer with useful information.

    1. SHG Post author

      As to your first point, and ignoring your hyperbole (information is not necessarily “insight,” and more information is often less insightful, but that’s not often appreciated), that’s fine, but that’s as a lawyer. No one has suggested there is nothing valuable to be learned, so you’re barking at a strawman. The problem is this being marketed to non-lawyers. That’s a different issue.

      As to your second point, that’s hardly what the analytics mean. You either didn’t read closely or didn’t grasp the implications. It sounds like a whole lot more, and therein lies the problem. Again, if this is only used by lawyers, then it’s not an issue. Lawyers can sort it out. But if it’s being sold to non-lawyers, then it’s a very real problem. And while it’s unclear whether it’s being marketed to non-lawyers, there is nothing that says they won’t sell it to anyone who asks.

      1. Adam Ziegler

        I looked closely at the sample report on the page to which you linked. The information shown entails: (1) factual docket information [docket #, venue, filing date] about patent cases involving the company sending the letter; (2) factual information about the lawyer/firm that sent the letter, its recent patent case filings and the clients it has represented in patent cases; (3) the historical litigation stats of the patents-at-issue [times asserted, open cases, findings (in a checkbox format), damages]; and (4) a listing of firms and companies that have most often been adverse to the company that sent the demand letter.

        To me, the last item is the most important. The implication I grasp is that it will make it easier for clients to find lawyers with relevant experience. Although I presume Lex Machina will monetize that “lead generation,” they deserve credit for building a report that’s designed to guide client-users toward qualified representation, not away from it.

        1. SHG Post author

          To me, the last item is the most important.

          To you. To someone disinclined to pay money to a lawyer, not so much. If we only look at things through our own prism, we miss seeing the problems that other eyes will see.

          One of the recurring challenges with the introduction of technology into law is that people love the upside and ignore the downside. They come together, and both need to be recognized.

      2. Patrick Maupin

        “information is not necessarily “insight,” and more information is often less insightful, but that’s not often appreciated”

        Au contraire, I think that’s usually highly appreciated. What are Lexis/Nexis, Westlaw, google, bing, etc., if they aren’t mechanisms that allow you to winnow down the data to items that might be relevant? Yet your point stands — for something complicated the best you can hope for is relevant result that can help you develop the insight.

        And yet, for the simple things, google actually does a good job with the insight. When I want to know when my wife will be back from England, I can type “AA55” and it will tell me.

        “And while it’s unclear whether it’s being marketed to non-lawyers, there is nothing that says they won’t sell it to anyone who asks.”

        I personally would welcome such a search service if it were reasonably priced. Will I be damaging myself by not contacting a lawyer first? Perhaps. How would the legal profession keep me from falling into this trap? Assuming the proper disclaimers are put into place, I’m not really sure how you could regulate it without running afoul of the First Amendment, because it really just becomes Nolo Press on steroids.

        1. SHG Post author

          Nolo Press, another danger by the way, it isn’t, at all. It gives qualitative advice, not merely quantitative, pick-a-form-and-screw-up-all-you-want. It gives you “odds,” pushing you toward one direction or another, with the intended impression that one will work while another will fail. It “advises.” Except it’s not true, but you won’t know that or why.

          And Google is only as good as your query, and even then, run through all 177,800,953 returns and let me know whether it narrowed it down sufficiently for you. But if you only look at the first page, maybe you got snookered by SEO cretins and don’t know it.

          1. Patrick Maupin

            > Nolo Press, another danger by the way, it isn’t, at all.

            Nolo Press has been in litigation that shows that bar concerns don’t trump first amendment concerns. And when you take into account that google has been in litigation that shows that machine results can also be afforded first amendment protection (for reasons that I can’t begin to describe nearly as well as Eugene Volokh), it seems to this layman that meaningful regulation of Deus Ex Machina might be a hard row to hoe, although I don’t discount the possibility they might be privately sued into oblivion if they give bad advice.

            > It gives you “odds”

            And if they don’t carefully explain that past performance is no guarantee of future results, they are idiots.

            > It “advises.” Except it’s not true, but you won’t know that or why.

            If they’re not showing me the raw data along with their analyses, I’m not interested.

            > And Google is only as good as your query

            No question about that, although I am still occasionally surprised when people I know don’t get good results out of it.

            1. SHG Post author

              You’re a smart guy, and I bet you would make good use of analytics. A lot of people aren’t very smart, and use cheap and easy as a facile substitute without understanding or appreciating the harm they can do to themselves.

              I don’t care about protecting turf. I care about people being harmed, and I care about preventing needless harm in the name of quick money. The issue isn’t speech, but holding oneself out as competent to provide legal advice. You can opine as to your views on the law all you want. That’s fine with me. You just can’t sell your services to provide legal advice unless you’re qualified to do so.

              That’s how speech works. That’s how protecting people from themselves works. We can have both.

            2. Patrick Maupin

              > That’s how speech works. That’s how protecting people from themselves works. We can have both.

              I would like to think so — current precedent says the law has the tools to deal with commercial speech that goes overboard. But these tools are often either blunt or mis-used — for example, the FDA allows outright lies in some cases (e.g. 0 g trans-fats), and suppresses fairly nuanced free speech in others (e.g. Diamond Walnuts). And while I believe you are sincere about not wanting to protect turf, I also have every reason to believe that some of your colleagues are not quite so benevolent. Of course, it’s arguable that Nolo can do what it does precisely because the lawyers who are on the side of free speech are in a better position to rein in the excesses of the bar than they are to rein in the excesses of the FDA, so I’ll just have to reserve judgment for now.

            3. SHG Post author

              My problem with NoLo isn’t speech, but putting a loaded gun in the hands of a child. They may be allowed, but it’s not going to end well.

Comments are closed.