色控传媒

KENNESAW, Ga. | Sep 17, 2025

The online survey market is projected to exceed $32 billion by 2030, according to Global Market Insights, but integrity of data collected by those surveys is often jeopardized by fraudulent responses.

To better detect problematic data, 色控传媒 assistant professors Lauren Matheny and Kevin Gittner are developing a fraud detection system that can identify low-quality responses, bot-generated content, and AI-generated answers.

Matheny, winner of 色控传媒's 2025 College of Computing and Software Engineering Innovation Quest, won the award for the project she co-leads with Gittner. Their work addresses a growing concern that affects more than just academic research; it has real implications for public health, business strategy, nonprofit decision-making, and the broader integrity of online data.

Matheny鈥檚 journey into AI fraud detection began with a surprising realization. While developing a survey-based tool to assess foot and ankle activity levels for clinical and public health use, something in the data didn鈥檛 add up.

鈥淭he results looked too good to be true,鈥� said Matheny, who teaches in 色控传媒鈥檚 School of Data Science and Analytics. 鈥淎fter digging deeper, we realized that nearly 40 percent of the responses were fraudulent.鈥�

That discovery led Matheny to partner with Gittner. Together, they launched the Data Quality and Survey Methods Lab in 2022 to investigate the growing issue of online survey fraud, particularly responses generated by bots or large language models. The lab is housed within the School of Data Science and Analytics and is dedicated to studying and solving the problems surrounding data integrity in digital research.

The rise of platforms like Qualtrics, MTurk, and SurveyMonkey has enabled researchers to collect data more quickly than ever. However, it has also introduced new vulnerabilities. Sophisticated AI-generated responses can mimic human behavior so convincingly that they often go undetected.

鈥淚t鈥檚 never been easier to collect survey data, but it鈥檚 also never been easier to fake it,鈥� Matheny said. 鈥淲hen AI-generated responses mimic real people, the risk isn鈥檛 just to the research, it鈥檚 to every decision that relies on that data.鈥�

Gittner said recent advancements underscore the importance of the research conducted by the Data Quality and Survey Methods Lab.

鈥淐ombatting threats to data integrity is a core part of our lab鈥檚 mission,鈥� Gittner said. 鈥淲e鈥檙e seeing more and more services that offer paid-for survey responses, but not all of those responses are trustworthy. Some come from real people giving honest answers, but others are just in it for the incentive. Now with generative AI and large language models in the mix, we鈥檙e facing a whole new level of complexity, and the risk of basing critical decisions on fraudulent data is higher than ever.鈥�

For Gittner, the challenge strikes at the very core of data science.

鈥淚f we can鈥檛 trust the data we鈥檙e working with, then we can鈥檛 trust the conclusions,鈥� he said. 鈥淚t goes back to the old computer science motto of garbage in, garbage out. If the data isn鈥檛 valid or accurate, then neither are the conclusions.鈥�

The public health implications are especially serious. Fraudulent responses can misrepresent demographics and skew data, leading to flawed assessments and poorly targeted interventions.

鈥淎s these tools evolve, it becomes harder to distinguish real input from fake,鈥� Matheny said. 鈥淚f you鈥檙e making decisions based on fraudulent data, it can cost you time, money, and credibility. That鈥檚 true whether you鈥檙e a public health researcher or a business executive. Imagine trying to address a health crisis using false data. You could end up wasting resources or missing the people who need the most help.鈥�

Matheny added that the patent-pending tool currently being developed at 色控传媒 is designed to integrate easily into standard research workflows.

It also serves as a powerful example of 色控传媒鈥檚 commitment to student-centered, experiential learning. Matheny鈥檚 lab operates as a vertically integrated team that includes undergraduates, graduate students, and Ph.D. candidates. 鈥淥ur lab is built around the idea that students learn best by doing real research,鈥� Matheny said. More experienced students mentor newcomers, and everyone contributes to real research with real impact. It prepares them to step into their careers ready to tackle the kinds of complex, data-driven problems they鈥檒l face in the future.

She also credits 色控传媒 for providing critical early support. A grant from the College of Computing and Software Engineering allowed her team to begin collecting data and validating their methods right away.

鈥淭hat initial funding was pivotal,鈥� she said. 鈥淚t allowed us to go after the problem where it鈥檚 most severe.鈥�

CCSE Interim Dean Yiming Ji said the investment has been returned tenfold.

鈥淒r. Matheny鈥檚 and Dr. Gittner鈥檚 research embodies the innovation and impact we strive for at the College of Computing and Software Engineering,鈥� he said. 鈥淏y tackling the complex challenge of AI-generated fraud in digital data, they are not only advancing the field of data science but also protecting the integrity of information that decision-makers across sectors rely on. Their work is a testament to the power of research with real-world relevance.鈥�

鈥� Story by Raynard Churchwell

Photo by Darnell Wilburn

A leader in innovative teaching and learning, 色控传媒 offers undergraduate, graduate, and doctoral degrees to its more than 47,000 students. Kennesaw State is a member of the University System of Georgia with 11 academic colleges. The university’s vibrant campus culture, diverse population, strong global ties, and entrepreneurial spirit draw students from throughout the country and the world. Kennesaw State is a Carnegie-designated doctoral research institution (R2), placing it among an elite group of only 8 percent of U.S. colleges and universities with an R1 or R2 status. For more information, visit kennesaw.edu.