Here’s a scenario that might be common enough in your day-to-day life: imagine that you’ve misplaced your wallet. (Apparently, Americans do this a lot. Statistics show we spend two and a half days each year looking for misplaced items.)

But fortunately, you know exactly what it looks like. You also know that it’s somewhere in the 400-square-foot hotel room you stayed in last night. It might take you a minute, but you should be pretty confident that you’ll walk out the door with your wallet before long.

Now, suppose you’ve been asked to look for someone else’s lost wallet, but you aren’t really sure what it looks like. It might be an exploding George Costanza wallet, a thin sleeve with a credit card or two, or even just a naked money clip. And your search isn’t confined to a hotel room; all you know is that this wallet is lost somewhere in Belize.

Oops.

When faced with those parameters, it’s not really a personal failing to admit that you’re not likely to find that wallet, right? Even if you somehow manage to succeed, combing through an entire country for an indeterminate object probably isn’t the best use of your time.

That’s what it’s like looking for discoverable information on the web: the information is there, but holy cow, it might be hard to find. To succeed, you have to know what to look for, where to look, and how to look—and even then, you might need a few years to get the job done.

When you’re up against an investigation that’s that vast and unconquerable, you need tools to help you out. That’s where the power of artificial intelligence (AI) comes to bear.

The Truth Is Out There

These days, life happens online. People search for their perfect home or their next job online. Heck, most people nowadays find their spouses online. We read the news online, keep up with our extended networks of family and friends through online social media platforms, buy a growing percentage of our groceries and consumer goods online, and get the answers to all of our most pressing, and private, questions at the tap of a keyboard.

You better believe there’s discoverable information to be found on the web.

Take the case of Robinson v. MGM Grand Detroit, LLC, No. 17-CV-13128 (E.D. Mich. Jan. 17, 2019). In this employment discrimination case, the plaintiff, a valet attendant at MGM Grand Detroit, claimed that he suffered discrimination because of both his age and his disability. But he refused to produce certain information about his activities during discovery. His employer fought back, obtaining records from his gym that indicated he was working out while he was on FMLA leave. With that support, the employer asked the court to order that the plaintiff produce additional information from his Facebook account, Google photos, and Google location data. The court agreed that this online evidence was relevant and proportional to the plaintiff’s claims and, thus, discoverable.

Or consider Papadakis v. Fresh Meadow Power NE LLC, No. 526239, 2018 NY Slip Op. 08728 (N.Y. App. Div. Dec. 20, 2018). There, the claimant received workers’ compensation after sustaining a head injury at work. He claimed that because of his injuries, he couldn’t “sit upright or stand without head support.” But someone got suspicious, and at some point, the claimant was put under surveillance for an alleged violation of the state’s workers’ compensation law. His former employer and its insurer produced video footage showing him “lifting and carrying wood, engaging in an activity involving a soccer ball with a child, presumably his son, taking items out of the rear of a vehicle, including a paintball rifle, and carrying firewood and placing it into a pit.” They also submitted photographs and location information from the claimant’s social media accounts, showing him “socializing and smiling at various locations, including a Disney park and the beach.” Again, oops. In this case, the court ruled that the claimant had, in fact, violated the law and disqualified him from receiving future wage-replacement benefits.

And then there’s Hinostroza v. Denny’s Inc., No. 2:17-cv-02561-RFB-NJK (D. Nev. June 29, 2018). Here, the plaintiff claimed she was injured in a slip-and-fall accident at a Denny’s restaurant. The defendant moved to compel information about the plaintiff’s two prior slip-and-fall accidents (hey, some of us are just clumsy) as well as a car accident she’d been in the year before. It also asked for information from the plaintiff’s social media, since she claimed damages for “emotional distress, adverse emotional or mental state, and other emotional injury.” The court agreed that “[social networking] content is neither privileged nor protected by any right of privacy.” Further, “social media activity, to an extent, is reflective of an individual’s contemporaneous emotions and mental state,” and is therefore relevant.

What can you learn from these cases? Until you look, you never know what you might find online that could help you establish a claim or defend against someone else’s allegations. Don’t just take it from us: personal injury attorneys everywhere advise prospective clients to be careful what they say on social media. One firm notes that “what seems harmless to the account holder can be made to look like much more by the defense lawyer.” This firm also points out that “your social media posts are the best proof of your past because they don’t depend on what other people are saying about you. They are your own words and or pictures or video of you … [which makes it] next to impossible to deny them.”

But how do you find the needle in the haystack of online information?

How Do You Find What You’re Looking For?

Well, there’s the rub: with all the information that exists online, how do you find something that could help with your case? How do you even identify the person involved if you don’t already know about their social media profiles? What name do they use online, and on what platforms? How much information do they share publicly? What chatrooms or discussion boards or dating sites do they participate in?

It not your fault that you literally can’t do this alone. That’s why artificial intelligence investigations are such a game-changer. With smart technology, you can point a computer toward the web—yes, the entire universe of online data—and ask it to find information about a particular person. Which it will then do, tirelessly, until it finds whatever there is to find.

These tools are already being used in other high-stakes investigations. Take insurance fraud, which is estimated to cost $80 billion a year. Insurance investigators now use AI to cross-reference social media profiles, aggregate data about specific users who may be of interest, and identify potential markers that could indicate fraudulent activity. For example, they might note disparities between claimed activities and online activity logs, differences in location data, or relationships between claimants.

That’s not all: financial institutions use AI to “identify anomalous transactions” that might indicate money laundering or other financial crimes. There’s even an AI tool created by researchers at the University of Southern California that “scours the web for digital breadcrumbs scattered in sex ads, allowing investigators to find victims and shut down sex rings” and other online sex traffickers.

There’s no limit to what you can find online, but it is helpful to divide and conquer. As the human, your job is to think creatively about what information might be helpful to your case (or harmful to your opponent’s case). The computer’s job is to sift through the enormous web of information available online and find the good stuff, or at least identify something that will lead to useful information. With AI, you can tighten up the definition of what you’re looking for and narrow the scope of your search from “Belize” to “this hotel room”—which gives you a much better shot at success.

It takes artificial intelligence to conquer the artificial—and overwhelmingly large—data source that is the web. Humans can’t beat machines at data, but we can work together to figure out what information we need and where we can find it.

[View source.]