IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

FETC25: Data Masking to Protect Schools from Data Theft

By creating “fake” versions of their data to interface with ed-tech vendors, school districts can limit what kinds of personally identifiable information those companies have in the event they experience a data breach.

Image of a person's face overlaid with numbers and data
iftistock
To some cybersecurity experts, the data breach of PowerSchool earlier this month fit an emerging pattern. It followed a similar breach of the online learning platform Learnosity six months prior, and the school security company Raptor Technologies six months before that. After years of going after school districts, cyber criminals are now targeting ed-tech vendors, too.

In a Wednesday session at the Future of Education Technology Conference in Orlando, Fla., ClassLink Sales Manager Mcdeny Alcantara explained why.

“I sadly lived through a ransomware attack on one of my school districts, and they wanted our money, but more than that they wanted our data. Because if they don’t get our money, and they have our data, they can resell that and do even more damage in the future,” he said. “If I [as a hypothetical bad actor] cannot get to your data, because you did a great job protecting your network, I may want to target your district partner, and they may have some vulnerabilities that we can exploit and take your data out. That’s what bad actors are doing. Not just education, but education partners.”
In response to this, Alcantara recommended that school districts start the practice of data masking.

Admittedly, he has a personal stake in this recommendation, as his employer helps schools to do this with a service called DataGuard. But he described the concept as a way for school districts to limit their liability if one of their vendors should be compromised in a cyber attack.

To begin with, Alcantara said, school districts should guard their data by having or creating a data governance program, prioritizing the most impactful security measures for their network, and documenting their security policies. He cited best practices that will be familiar to many IT security professionals already: single sign-on with multifactor authentication, data encryption and application vetting.

He recommended adding to that list the practice of data masking, which creates “fake” versions of data on a school district’s roster, which is then stored on a separate server that interfaces with vendors.

“You have the restore system — it could be your [student information system] — feeding your roster server, and your roster server is the one that interfaces with your ed-tech vendors,” Alcantara said. “The roster server is the one that you control, and you decide what to mask, what not to mask, to whom and when. So, your source system data is never altered. What you share with vendors is what you decide, how much you give and how much you don’t give.”

Rachelle Corry, coordinator of digital resources at Deer Park Independent School District in Texas, praised the practice of data masking as a customizable way to make data legible for users but not for external parties.

For Deer Park ISD, Corry chose to mask data on the student roster so that only the first three letters of a student’s first and last name were visible, followed by letters and numbers — for example, “Nic9c6f6 Brad4530” instead of the student’s full name. Teachers said they could still identify which students the masked versions of the data belonged to, but if the ed-tech vendor were to have its data stolen, the information would be useless to the thieves.

Corry said this can be effective as long as it’s communicated clearly to teachers, and they can give feedback.

“In our case, I didn’t want the teachers to be confused, or handicapped, or not being able to function,” she said. “When I first started, I reached out to some teachers and said, ‘I think you’re going to see some changes, and I need some feedback,’ because we don’t want to rock their world. They’re stressed enough in the classroom already.”

Alcantara said once data is masked, outside parties cannot reverse-engineer or revert it back to its original values unless they have access to the original data set on the school's network. They can't get it from the vendor.

He added that the customizability of the function is also key.

“It’s not one-size-fits-all,” he said. “You can apply this to some, many or none of your applications.”
Andrew Westrope is managing editor of the Center for Digital Education. Before that, he was a staff writer for Government Technology, and previously was a reporter and editor at community newspapers. He has a bachelor’s degree in physiology from Michigan State University and lives in Northern California.