Social monitoring started in the world of marketing, allowing companies to track what people were saying about their brands. But now, with software that allows users to scan huge volumes of public postings on social media, police are starting to embrace it as well.
Many police departments in Britain use a product sold by CrowdControlHQ. CEO James Leavesley calls it a "social media risk media and monitoring" company, meant primarily as a means of staying in touch with the public. But Leavesley says it's also a way to detect trouble.
"By looking at keywords, it can track conversations," he says. "Vulnerable people have been identified and reached out to."
In the U.S., a company called BrightPlanet sells a Twitter-monitoring service that is more explicitly marketed as an investigative tool.
"If you had 1,500 gang members, like we do in Detroit — we have their handles, so we're able to identify what the gang members are doing," says BrightPlanet vice president Tyson Johnson.
The tool, called BlueJay, is capable of scanning the entire "firehose" of tweets, he says — far more than is available to search from the Twitter web page. It can be configured to focus on tweets coming from certain places, and it can collect instant photographic evidence from a disturbance.
"If we'd been able to monitor real-time during Boston Marathon [bombing], they'd have an immediate repository to interrogate, as soon as the bombs happened," Johnson says.
Location information depends on people leaving their geo-tagging option on, and only a small subset of Twitter users do. But when they do, BlueJay can track their movements over time on its map. For police, it's a potential goldmine of information.
"It's like a stake-out," says Vernon Keenan, director of the Georgia Bureau of Investigation. "If persons are talking about certain criminal activity, it alerts us to it."
But some police hesitate to use these tools. Keenan says understands that this is all public information — that people tweet this information voluntarily — but that there are many public things that the police should not monitor. He gives the example of a political rally.
"For law enforcement to be there and to take photographs of all the participants — monitoring — is not against the law, but it's not acceptable," Keenan says.
So even on the public Internet, Keenan requires his agents to get permission from a supervisor before they scan social media. They have to explain what they're monitoring and why.
Ryan Calo, a professor at the University of Washington law school who specializes in privacy issues, says police could run into trouble with searching on the Internet.
"If officers were [scanning social media] on the basis of gender and then making decisions on that basis, you could run into constitutional scrutiny," Calo says. "And you'd be almost sure to if your keyword involved the word 'Muslim.'"
Calo says the law is fuzzier when it comes to other kinds of searches, such as political keywords. The law and the courts are far behind the technology, and no police department wants to become the test case. Calo says it's not clear whether it would be illegal for police to monitor for a keyword such as #Occupy, but that doesn't mean police should feel free to do so.
"Any police officer ought to sort of think through a kind of publicity principle, which is, 'If it were to get out that we did this exact search, what would the public reaction be,'" says Calo.
That's why Keenan is now campaigning to get more police departments to set up internal rules for social media scanning. He thinks the tools are useful, and he's worried that a public backlash could cause law enforcement to lose them.
At the annual meeting of the International Association of Chiefs of Police, Keenan gave a speech warning his colleagues that social media monitoring is a "hot stove issue" for police.
"[And I] know what happens when you touch a hot stove — you get burned," he said.