The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
I'm blogging excerpts this week and next week from a new article by Mark Lemley and me, "Law, Virtual Reality, and Augmented Reality." (Click on the link to see the whole article, including footnotes.) This post is from the criminal law part; we begin with disturbing the peace, but soon we'll tell you about indecent exposure, strobing, virtual groping and more.
One general assumption we make throughout: VR and AR technology (think something like Google Glass) will get better, cheaper and more effective at rendering lifelike human avatars that track the user's facial expressions. I think that's a safe assumption, given the general trends in computer technology, and one that doesn't require any major technological breakthroughs; but we are indeed prognosticating about technology that's likely at least a couple of years out.
Much traditional criminal law enforcement involves street crimes: in-person misconduct, such as robbery, sexual assault, indecent exposure, or disorderly conduct. Many such crimes literally happen on the street. Many others happen in homes, businesses, or schools, but share many traits with traditional street crimes.
Many of the worst such crimes aren't a problem in VR. You generally needn't worry about being really murdered in a virtual space. Likewise, you needn't worry (subject to some complexities that we'll mention below) about being really beaten or raped.
Indeed, this could be one reason people will shift some activities to VR. Physically going out to drink with friends might be more fun in some ways than getting a virtual drink, where everyone is physically at home but can see each other in VR. You can hug your friends in a real bar. You can feel physically close to them and not just emotionally close. If you're looking to pick up a sex partner for the evening, doing that in VR would require haptic hardware that goes beyond what we have today.
Yet going out together for a virtual drink—to be precise, staying in for a drink, but being virtually together—has its own advantages. You needn't worry about getting into a bar fight, or getting mugged on the way home. You needn't worry about driving home drunk, or paying for a cab. Plus, the booze is much cheaper at home.
Still, as we'll discuss below, there may well be some kinds of "street crime" in VR. How will the law likely deal with that? How should it?
1. Disturbing the peace and the Bangladesh Problem
What sorts of street crime can there even be in VR? Today's VR is basically audiovisual—you can see and be seen and hear and be heard, but you can't be punched or shot or caressed. (Caressed is surely on its way, but not here yet.) We thus focus on crimes of sound or of sight.
A classic sound crime is disturbing the peace through loud noise, for instance through screamingly loudly in a public place. That crime can pose First Amendment problems when applied to speech that disturbs because of its content, but it's pretty straightforward when applied to speech that disturbs because it's too loud.
Indeed, if you see someone standing on the sidewalk screaming, calling the police is a standard response. You expect the police to come out, maybe talk the guy into going away, maybe arrest him, maybe even have him prosecuted. Dealing with such annoying street behavior is part of what police normally do.
Now say someone is screaming in a VR public place. Let's assume this isn't in a game, but in a place where people need to congregate for economic reasons—to shop at a VR store, or even go to their VR jobs. The harm caused by the screaming is the same: It interferes with people's other tasks.
So you call the police.
"Officer, there's this guy screaming and bothering my kids and me."
"What's the street address?"
"It's not on the street, it's in this VR world."
"We're playing a virtual game in the virtual park, and this guy is bothering us."
"Where are you, really?"
"Well, I'm sitting in my bedroom, but that's not what's important! I'm wearing my virtual headset, and it feels to me like I'm playing with my kids in the park—they're with my ex across the country, but we're spending some time playing together, and this jackass is ruining it for us."
"And where is he, really?"
"Oh, I clicked on his avatar, and it tells me that he's hooked up from Dhaka—you know, in Bangladesh. But it feels like he's right next to us."
Now maybe if you call a more technically savvy police agency, they'll understand your concerns more quickly. But their reaction is likely to continue to be skeptical, because of what we label the "Bangladesh problem": It will take a lot to get domestic police interested in investigating a crime where the criminal is in a foreign country. Indeed, it will take a lot even if the criminal is in another American state, or perhaps even in another city. Getting some extradited is a hassle. Even dealing with another jurisdiction's police department to arrange an arrest in the same state is a hassle.
Will they go through the hassle to investigate a murder? Maybe. But, "You think I can get someone extradited from Bangladesh for disturbing the peace?," the police officer might ask you. "Or even from Nebraska?" Indeed, perhaps your state won't even have jurisdiction over such crimes committed by people screaming in their rooms elsewhere in the world; but even if the state is legally entitled to prosecute such crimes, it would surely be very hard for local police and prosecutors to bring such a prosecution.
And there's every reason to think that the VR street criminals would indeed live all over the world. There are no oceans or borders in VR—that is one of its advantages. The VR "places" in which Americans will travel will be disproportionately Anglophone (though good real-time translation might change that), and disproportionately drawn from richer countries. Yet many of the people who share the same VR "street" will be oceans apart, and most will at least be from different states.
The same problem already exists to a significant extent on the Internet. The people who harass you or even threaten you on Twitter or Reddit, can as easily be in South Africa as in South Carolina. Courts handling civil cases have struggled for decades with how to address the problem of people who cause injury far from where they live. But criminal prosecutions for such transnational threats appear to be vanishingly rare.
Yet the illusion of presence that VR and AR bring is likely to make potentially criminal incidents more common. It's relatively rare for someone in a foreign country to care so much about us that he would tweet death threats about us; it happens, but only for pretty high-profile people. Most threats seem likely to stem from personal, emotionally laden interactions that usually require a sense of in-person connection—people threatening their exes, gang members, schools, and the like.
But the crimes we describe in this subsection and the coming ones are likely to be much more common. People scream and create a public commotion in the real world; there's no reason why they wouldn't do the same in a VR space. People indecently expose themselves in the real world; there's no reason why they wouldn't do the same in VR (more on that below). Indeed, they may be more likely to do this, precisely because they may reasonably infer that it will be hard for the police to catch them. Moreover, there will be more desire for criminal prosecution than with Internet misconduct, precisely because the feeling of physical presence may make the victims of VR street crime viscerally feel victimized. That desire, though, may be hard to satisfy.
To be sure, VR does tend to facilitate policing in one way, by solving some problems of proof and identification. If the VR platform keeps good logs, it can accurately report just which avatar was screaming, and just how loud he was. And perhaps the VR platform requires people to identify themselves before accessing it, at least with a credit card; or with the proper subpoenas, the typical avatar can be traced back to an Internet subscriber. But the greater difficulties caused by extradition are likely to exceed the greater ease of proof. And many VR street crimes might thus be practically ignored by traditional police department.
Of course, this might yield pressure for VR operators to set up in-VR "police," who might be able to deal with transgressors quickly; and there might be "courts" as well, for resolving disputes (especially disputes involving in-VR commerce). But the penalties will likely be, at most, suspension or ejection from the VR environment. And it seems likely that the ejected participants can just get back on by creating a new user ID.
If a VR environment requires people to provide a credit card, or otherwise supply a deposit, such new user IDs might become harder to create, and the environment might even threaten fines or forfeited deposits for bad behavior. How often this will happen will depend on economic factors that we can't easily predict. We expect that many VR environments will want to allow free access, or at least access that doesn't require a credit card (but might require only some prepaid gift card), since the VR operators will want to harness network effects by increasing their user bases. Presumably, those operators will make money from in-VR purchases rather than through credit card subscriptions. But we're not certain whether this will be so; indeed, some environments might want to require credit cards or elaborate identification systems precisely to maintain a more orderly experience for their users.
So the real-world police are unlikely to intervene to stop the VR street screamer. But there's a good reason why disturbing the peace is a crime: It affects people's quality of life, and tends to push them away from a place where they want to be, and where we might want them to be (for instance, if we want them to work there or shop there). And the creators of the VR environment will be keenly aware of this, because lost quality of VR life means lost profits to them, especially since different VR environments will likely be hotly competing with each other.
Code, as Larry Lessig put it, is law—maybe the most effective sort of law. And VR environment operators can easily implement code that can deal with the screamers. The operator could, for instance, allow each user to control the perceived volume, for that user, of any other user. That's good not just to silence the screamers, but also to quiet down acquaintances who are a bit too loud, or to amplify acquaintances who mutter. And this should be technically trivial to code.
The instruments of the real world—real ears and real brains—don't have such a feature. But the sensescape created by the VR software is more versatile and more individually controllable than what mere human anatomy can provide. [Footnote: By "sensescape," we simply mean the array of sensory inputs that a VR environments provides to users: today, mostly sights and sounds, but it could soon include touch, smell, temperature, pain, and more.] Taking advantage of this versatility can help prevent or quickly interrupt VR street crime. Yet shifting to these in-VR remedies likely means shifting away from the criminal law, and from the standard criminal law penalties.