Classroom

What Can Universities Do About Contract Cheating?

Academic cheating sites are on the rise. Here’s how universities are detecting contract cheating and unauthorized document sharing.

Amelia Pang

Amelia Pang is a journalist and an editor at EdTech: Focus on Higher Education. Her work has appeared in the New Republic, Mother Jones, and The New York Times Sunday Review, among other publications.

EDITOR’S NOTE: This is part 2 of a 2-part series on contract cheating. In part 1, EdTech went undercover to see what students post on academic cheating sites.

The shadow industry of contract cheating — where students hire a third party from the internet to do their coursework — is not a new phenomenon. But the number of students who turn to such services, as well the number of websites that offer them, have rapidly increased during the pandemic. And even cheaper artificial intelligence bots that can produce essays are gaining traction among desperate students.

An EdTech investigation found that some academic cheating sites have invested in automation technologies that can efficiently match students to an academic ghostwriter with the expertise to cover that topic. To protect themselves from legal trouble, some homework help sites can also automatically add lines such as, “I need a sample draft to help me study,” to student requests that specifically state they are asking for a homework ghostwriter to complete the assignment for them.

Although there is no magic pill for curing the cheating epidemic, it is clear that higher education institutions need technologically advanced strategies to tackle tech-savvy cheating sites like college paper writing services. That’s why in recent years, a growing number of academic integrity departments have been partnering with IT teams.

“There is some overlap with what our cybersecurity teams do,” says Evan Lowry, assistant general counsel at Southern New Hampshire University. “For example, detecting a contract cheater who is entering the learning management system is not so different from detecting other types of intrusions. You’re looking for anomalous behaviors, like multiple login attempts. And maybe an impossible login scenario, like a student logging in from California five minutes ago, and then logging back in from the other side of the world.”

As higher education IT departments rise to the challenge of protecting academic integrity in online classrooms, here’s a look at what some universities and colleges are doing.

Crawler Bots Seize Course Materials from Cheating Sites

At the University of Maryland Global Campus, IT innovation has played a pivotal role in helping the institution detect and deter contract cheaters at scale.

It came down to a battle of the bots.

UMGC has developed an AI-enabled crawler bot that can scan hundreds of academic cheating sites at once.

“UMGC uses machine learning technology to vacuum up our materials from over 700 sites that are the most common offenders,” says Douglas Harrison, vice president and dean of the school of cybersecurity and information technology at UMGC.

Detecting a contract cheater who is entering the learning management system is not so different from detecting other types of intrusions.”

Evan Lowry assistant general counsel, Southern New Hampshire University

The machine learning technology allows UMGC to efficiently detect flagged content and compare it with a database of the university’s assignments. Whenever there is a match, the bot automatically sends legal takedown notices to the offending sites under the Digital Millennium Copyright Act, the legal avenue that institutions use to reclaim copyrighted course materials from academic cheating sites.

The bot is designed to check back in 30 days to ensure the flagged materials were indeed removed from sites. “All of this is automated, and it allows us to respond to the circulation of our documents without authorization at a scale that no human being could ever do,” Harrison says. “It’s a massive game of whack-a-mole with these sites. You have to have technology that’s intervening at the scale of the threat.”

COMBAT CHEATING: Learn how to implement automation solutions that can augment your academic integrity department.

Computer Science Students Catch Online Learning Cheaters

Meanwhile, professors at Embry-Riddle Aeronautical University worked with computer science students to develop a similar search engine tool called CourseVillain.

Educators at Embry-Riddle can use the CourseVillain search engine to hunt for university content on a website that faculty believe students frequently misuse to engage in academic misconduct. The tool makes it easier for faculty to detect coursework that has been illegally shared, and automatically file copyright infringements when the institution’s documents are detected.

As of early April, Inside Higher Ed reported the CourseVillain tool had flagged 237, 293 documents as Embry-Riddle property on the site.

Algorithms Detect Major Deviations in Writing Styles

In recent years, breakthroughs in machine learning have also greatly advanced academic plagiarism detection algorithms.

UMGC is in the early stages of using an authorship authentication tool from a third-party vendor. The tool uses AI-powered algorithmic analysis of student writing to flag when a student’s writing style undergoes dramatic changes.

The algorithm analyzes the linguistic styles and semantic features of students’ writing samples, while also taking into consideration that student writing styles will likely improve over the semester. “We do want to see some change over time,” Harrison says. “But when we see a statistically significant variation, we have a human investigator look into that.”

Biometric Tech Verifies Identities of Exam Takers

In recent years, there has been a rise in students paying academic ghostwriters to take online courses and exams for them. This allows homework ghostwriters to access their classmates’ contact information and create fake study groups that lure desperate students into contract cheating.

To address this phenomenon, UMGC is exploring potential adoption of biometric identity authentication solutions. They are not the first to do so. Federal and state governments, for example, are already using biometric authentication to verify user identity. Technologies that create unique, biometric-based identity profiles of each student can help universities authenticate identities and continually reaffirm that the person logging in to the course is indeed the same person who registered for it.

It’s a massive game of whack-a-mole with these sites. You have to have technology that’s intervening at the scale of the threat.”

Douglas Harrison vice president and dean of the school of cybersecurity and information technology, UMGC.

This approach would mean analyzing the biometric signals that a student sends when he or she interacts with keyboards and devices — and pairing those signals with information about the student’s usual IP addresses to create a signature profile of what authentic, authorized use looks like.

“So anytime the system detects a deviation from the signature, that’s a signal for us to say, ‘We need to look into this a little further,’” Harrison says. “It would allow us to authenticate identity behind what single sign-on security software can do.”

With that said, Harrison emphasizes it’s important to assume a student is innocent until proven guilty.

“The technology gives us points of departure to explore what may or may not be a problem,” he says. “We have students worldwide. And we have many students who are deployed in the military, so their access points could look very strange on the backside of the IP address analysis. That doesn’t necessarily mean they’re cheating.”

“The goal of using these technologies isn’t to turn our students into adversaries, but to give us tools that help us protect the integrity of our resources,” Harrison adds.

Higher Education IT Leaders Win Institutional Support

Ever since UMGC embraced online learning over 20 years ago, the institution has been thinking about how to approach academic integrity in a digital environment.

They have tried myriad approaches, from using proctoring software to requiring students to take exams at a physical test center. But these strategies were not always effective, and did not ultimately align with the institution’s mission. “We moved away from proctoring over five years ago and closed all of our testing centers,” Harrison says. “We realized the best way to ensure authentic evaluation of learning is to create authentic assessments — not just have people do multiple-choice exams.”

UMGC focuses on project-based learning that uses real-world scenarios in which students encounter the kind of actions and operations they’ll engage in as part of their future careers. “The learning deliverable looks like the job that students will be doing in real life,” Harrison says.

“First of all, it’s a lot harder to cheat on authentic assessment,” Harrison says. “And second, we hope it means students don’t want to cheat on these assignments because they’re actually doing something that matters to them.”

After dedicating resources to designing better online courses and exams, UMGC also launched the Project to Combat Corruption and Attack. This initiative, which began three years ago, won them the funding to acquire and develop bots, as well as other authentication technologies in house. The university was able to accomplish this because it had already invested in robust IT infrastructure.

For other IT departments that are thinking of adopting a similar approach, Harrison says his advice is to get executive leadership onboard.

“We had presidential level support, and that’s critical for these things to work well,” he says. “It usually doesn’t take much work for executive leadership to endorse this type of initiative, because it is obvious how important this investment is for academic quality.”

momcilog/Getty Images