Three UMSI researchers earn Google Academic Research Awards to combat AI scams

University of Michigan School of Information researchers are working to design new protections to help people stay safe as digital scams become increasingly sophisticated. 

UMSI researchers Tawanna Dillahunt, Mustafa Naseem Florian Schaub have earned 2026 Google Academic Research Awards. Their projects will advance research on scam prevention and digital safety by creating meaningful interventions that prevent harm. 

Naseem and Dillahunt’s project, “Building Community‑Driven Scam Prevention: A Survivor‑Centered Approach to Digital Safety,” will expand on their ongoing community-engaged research by developing community-led strategies to avoid scams and strengthen digital resilience among Afghan communities in Michigan. 

The work builds on the researchers’ Community Tech Workers initiative in which trusted community members support neighbors with digital tools and services. During their work with public housing residents in Detroit and refugee communities in Washtenaw County, Dillahunt and Naseem noticed how refugee communities were being targeted by increasingly sophisticated scams. 

“We were starting to see that scammers have become more advanced and are microtargeting communities,” Naseem says. “For refugees, scams often involve impersonating government agencies or creating fake job opportunities to collect personal information.”  

Their project will focus on understanding how scams operate in refugee contexts and developing interventions that help communities respond. They will study scam patterns and vulnerabilities and co-design training materials with community tech workers to help people detect scams and intervene early. 

“Our goal with this particular grant is to try to understand how scams operate in a refugee context,” Naseem says. “Because there's very little research on that.” 

Financial losses from scams have disproportionate consequences for vulnerable communities. In some cases, victims don’t have savings to fall back on. Additionally, designing for these communities often produces solutions that benefit broader populations. 

“ I think as designers, if we design for the most vulnerable, we often end up with more resilient interventions,” Dillahunt says. “Interventions that work well for more people.” 

A central part of their project is building what Dillahunt and Naseem describe as “human infrastructure.” These are people who are in the community, know the culture and language, and can help others navigate digital systems and challenges. 

“In the past, communities had support networks like libraries, churches or neighborhood groups,” Naseem said. “As society becomes more digital, we need similar human support structures online. Instead of building another software tool, we’re building networks that create collective resilience.”

Schaub’s project, “User-Centric Deepfake Scam Warnings Nudging Verification Actions” will address the growing threat of AI-generated impersonation scams, known as deepfakes, which mimic a person’s appearance or voice. 

“There’s been a lot of research on helping people detect deepfakes,” Schaub says. “But given how quickly the technology is advancing, the ability for humans to reliably distinguish real from fake will become very limited.” 

Schaub’s work will explore how systems can effectively intervene when risks are detected. These can include pop-up warnings that users can act on during calls. 

“We’re investigating what meaningful verification actions people can take,” Schaub says. “For example, asking a question only a real person would know the answer to, or sending a text message and asking them to read it aloud.” 

Across both projects, the researchers emphasize that awareness alone is not enough to address modern scams. Effective solutions require both technological innovation and community support systems.

“There’s been too much emphasis on telling people to be careful,” Schaub said. “We need interventions that actually prevent people from falling victim.”

RELATED 

Learn more about UMSI researchers Tawanna Dillahunt, Mustafa Naseem and Florian Schaub by visiting their UMSI faculty profiles. 

Read more about UMSI research on privacy and security.