Bent over their computers in Salt Lake City, a dozen cyber sleuths scan the vast reaches of the Internet for contraband. Only, they’re not hunting traffickers of drugs or sex.
It’s standardized testing season across the U.S. — and they’re on the lookout for student tweets about the tests.
The web patrol team works for Caveon, a test security company charged with protecting the integrity of new Common Core exams developed by the publishing giant Pearson. To that end, they’re monitoring social media for any leaks about
test questions. News of the
surveillance broke this week, sparking a firestorm.
The American Federation of Teachers even circulated a petition demanding that Pearson “stop spying on our kids.”
But Pearson is hardly the only company keeping a watchful eye on students.
School districts and colleges across the nation are hiring private companies to monitor students’ online activity, down to individual keystrokes, to scan their emails for objectionable content and to scrutinize their public posts on Twitter, Facebook, Vine, Instagram and other popular sites. The surveillance services will send principals text-message alerts if a student types a suspicious phrase or surfs to a web site that raises red flags.
A dozen states have tried to limit cyber snooping by banning either colleges or K-12 schools, or both, from requesting student user names and passwords, which could be used to pry open social media accounts protected by privacy settings. Among those taking action: California, Illinois, Michigan and Utah.
At least five other states, among them New York and Maryland, are considering similar laws this session, according to the National Conference of State Legislatures.
But such laws protect only accounts marked as private. Many kids post publicly to build up their online followings.
And when they do, companies with names like Social Sentinel, Geo Listening, Varsity Monitor and UDiligence are there to read them.
The rise of online student monitoring comes at a time of rising parent protests against other forms of digital surveillance — namely, the vast quantities of data that technology companies collect on kids as they click through online textbooks, games and homework. Companies providing those online resources can collect millions of unique data points on a child in a single day. Much of that information is not protected by federal privacy law.
School administrators don’t tend to be too interested in that data, because it’s far too granular for them to make sense of it until an ed-tech company mines it for patterns.
But some principals, coaches and college presidents are acutely interested in student tweets.
Enter the surveillance services, which promise to scan student posts around the clock and flag anything that hints at bullying, violence or depression. The services will also flag any post that could tarnish the reputation of either the student or the educational institution. They’ll even alert administrators to garden-variety teenage hijinks, like a group of kids making plans to skateboard on school property.
Some of the monitoring software on the market can track and log every keystroke a student makes while using a school computer in any location, including at home. Principals can request text alerts if kids type in words like “guns” or “drugs,” or browse websites about anorexia or suicide. They can even order up reports identifying which students fritter away hours on Facebook and which buckle down to homework right after dinner.
Other programs scan all student emails, text messages and documents sent on a school’s online platform and alert school administrators — or law enforcement — to any that sound inappropriate.
The more comprehensive services attempt to break down anonymity offered by sites like Twitter and YouTube, where students don’t have to display their real names. Analysts cross-reference photos, map friend networks and even try to deduce class schedules from the timing of social media posts in order to unmask students who use pseudonyms online.
Sometimes the monitoring is covert: One company advertises that its surveillance software, known as CompuGuardian, can run on “stealth mode.” At the other extreme, some high schools and colleges explicitly warn students that they are being watched and advise them not to cling to “a false sense of security about your rights to freedom of speech.”
During standardized tests, the monitoring kicks up a notch. Several major assessment developers, including ETS, ACT and Smarter Balanced, told POLITICO that they do exactly what Pearson hired Caveon to do: Scan the web for any breaches in security, such as a photo of a test question or a Facebook post describing an essay prompt.
State officials often conduct their own sweeps of the Internet as well during testing season.
But it’s the day-to-day posts that those in the business of student surveillance find most illuminating.
Chris Frydrych, the CEO of Geo Listening, says his service routinely alerts school principals to students whose posts indicate they’re feeling particularly stressed or angry. He also points administrators to students who share too much personal information online, leaving them vulnerable to cyber predators.
Boasts about cheating. Dares to act recklessly. Taunts. Threats. Trash talk about teachers. For $7,500 per school per year, his service will scoop it all up and report it all to administrators.
“Our philosophy is, if someone in China can type in your child’s user name and see what they’re posting publicly on social media, shouldn’t the people who are the trusted in adults in a child’s life see that information?” Frydrych said.
He responds to critics who worry about privacy violations by quoting a student tweet he spotted while monitoring a school: “Twitter is not your diary. Get over it.”
Many education officials are uneasy with such services. “It’s not considered politically correct,” said Barbara-Jane Paris, past president of the National Association of Secondary School Principals. “It upsets students and stakeholders because they feel like they’re being spied on.”
But to some school administrators, cyber surveillance seems a natural extension of the age-old practice of deploying adult monitors to keep kids in line during lunch and recess. (Social Sentinel plays on that sentiment with its slogan: “Don’t leave the digital playground unattended.”)
“All the information is out there,” said David Jones, the president of Safe Outlet Corp., which markets the CompuGuardian software. The key, he said, is helping harried administrators sort through the enormously rich — and astoundingly intimate — digital trails each student unknowingly creates.
Is the principal concerned about kids goofing off in class? He can look at every program every student has installed on school laptops. Or he can track their web searches. Is he worried about bullying? He can snoop on online chats — reading both the student’s side of the conversation and any responses. Principals can even use the surveillance tools to figure out if they’re wasting money on apps no one uses.
“Everyone would just like for this to go away, but it’s not going to,” said Dan Domenech, executive director of AASA | The School Superintendents Association. “All the kids are on some form of social media now. You can’t ignore it.”
In the affluent suburban community of Deerfield, Ill., Superintendent Michael Lubelfeld signed up for a service called Gaggle, which scans student email and attachments, text messages, discussion boards and documents for words or images deemed inappropriate. The company says it discovers millions of inappropriate words and sends thousands of alerts to school administrators each year.
In Deerfield, Gaggle has unearthed just one serious incident in the past the 18 months — an eighth-grader emailing a nude photo of herself, Lubelfeld said. The student was given counseling. Gaggle has also notified Lubelfeld about a handful of less alarming incidents, such as the time a student tried to email a friend a photo of her baby brother with his naked bottom visible.
Lubelfeld also uses separate surveillance system, Hapara, that lets teachers see exactly which web pages kids have open when they’re using the school’s computers. The system, which costs $6 per student per year, also lets teachers take screen shots of a student’s computer or remotely shut down an inappropriate site, such as a video game. That system, too, has turned up few violations.
If nothing else, Lubelfeld said, it tends to keep the students in line: “Once the children realized that the teachers could use Hapara, they’ve pretty much stayed on task,” he said.
Other administrators, however, shy away from surveillance.
Hundreds of students follow the Twitter account of Jason Markey, the principal at East Leyden High School in Franklin Park, Ill. But Markey tells them all he will never follow them. If someone calls his attention to a student tweet that is offensive or frightening or abusive, he’ll absolutely take action, Markey said. Yet he most emphatically does not want to be in the business of cyber stalking his students.
“They need to know there is a separation,” Markey said. “Just because there are [surveillance] tools out there doesn’t mean that I have to look closer at what they’re doing online. I want to give them their space.”
Pearson’s monitoring program, first disclosed by New Jersey blogger Bob Braun, sparked anger from parents and teachers in part because the company has become a symbol of the controversial changes sweeping through public schools, such as the emphasis on high-stakes testing and the shift to the Common Core academic standards.
Irate parents and teachers have peppered social media with attacks on Pearson, many of them using the hashtag #Pearsoniswatching. The American Federation of Teachers has joined in, circulating a petition that features a creepy photo of a man in trench coat, fedora and sunglasses staring out a car window. The first line of the petition: “Big Brother really is watching.”
Pearson’s practice is not at all unusual. Only one testing vendor contacted by POLITICO, the American Institutes for Research, said it does not monitor social media. All others said they routinely scan such sites to look for breaches in test security.
They said they examine only public posts, not posts hidden behind privacy walls. And they said they use public sources, such as a Facebook account linked to a Twitter site, to verify students’ identity. If an inappropriate post is spotted, the companies generally contact state education officials, who reach out to the student’s school district to seek help taking down the post.
In a statement, Pearson adamantly rejected the notion that it is spying on students. “Absolutely not,” the company said.
In response to parent concerns, states using Pearson’s new PARCC exam did ask the company to stop cross-checking the names of students suspected of making inappropriate posts against the company’s list of registered test-takers. And New Jersey officials said Thursday that they would review the monitoring process to make sure student privacy is not compromised.
Yet state officials defended the basic principle of monitoring the web to ensure test questions aren’t leaked.
“The accounts Pearson is looking at are, by definition, public accounts with no expectations of privacy,” said Jacqueline Reis, a spokesperson for the Massachusetts Department of Elementary and Secondary Education.
