UM-Flint professors awarded $300K NSF grant for implementing cloud-based AI labs

Two computer science professors at the University of Michigan-Flint's College of Innovation & Technology have secured a $300,000 National Science Foundation grant to broaden student access to cybersecurity education.

The two-year grant will enable Suleyman Uludag and Halil Bisgin to expand on their work converting peer-reviewed research papers on artificial intelligence usage in cybersecurity into containerized Docker labs, which are application virtualization environments where students can practice bundling and running software in a way that works the same on any computer. The research papers can then be accessed through the National Artificial Intelligence Research Resource platform. The initiative democratizes cybersecurity research, removing financial barriers for students who otherwise could not afford high-powered computer equipment.

Unlike traditional cybersecurity labs, these exercises integrate AI techniques with real-world security problems, exposing students to actual research processes used in the field. Uludag and Bisgin have already piloted three labs with more than 150 students throughout five courses; they plan to make the platform available nationwide, incorporating added gamification elements that pit "red teams" of attackers against "blue teams" of defenders.

Setting up traditional labs consumes significant time for students and instructors, with frequent technical inquiries creating what Uludag described as "a headache" for everyone involved. But thanks to Docker containerization technology, creating self-contained learning environments is far more streamlined.

"Everything is already set up for you," Bisgin said. "The only thing that the student has to do is get the Docker image and put it on their computer, whether at home or in class."

Another major differentiator is the labs' foundation in published research. Each exercise walks students through replicating an actual peer-reviewed study, exposing them to authentic research methodology.

"They are basically reproducing the same thing that has been published in the literature by completing one of the labs," Bisgin said. "So it is not only a teaching tool, but also a gate to research."

The labs cover emerging topics at the intersection of cybersecurity and AI. One recent lab focused on password generation, in which students used large language models to create and crack passwords, exemplifying the gamification component planned for the grant period.

According to Uludag, NSF acknowledged that no comparable set of labs exists, setting the table for a truly novel approach to learning technology.

"There are few labs in existence that blend AI and cybersecurity as effectively as this one," he said. "And no learning platforms currently exist that can convert a published research paper into a teaching lab."

The grant, which began in September and extends to August 2027, will fund several initiatives beyond lab development. The professors plan to deliver tutorials at multiple conferences, organize an NSF workshop on campus, and hire undergraduate and graduate student assistants.

They also intend to extend the labs to high school students through summer camps, building on their previous experience mentoring young researchers.

"We are really trying to introduce this research discipline and research process paradigm at an early age," Bisgin said.

The project aligns with broader NSF initiatives. In August, the foundation announced a new funding opportunity of up to $100 million to support a network of programmable cloud laboratories—exactly the model Uludag and Bisgin have pioneered.

The professors have already published one paper on their methodology at an Institute of Electrical and Electronics Engineers conference, with another under review and several more in the pipeline. As the grant launches this fall, they hope their approach will provide a replicable model for bringing cybersecurity research into every classroom, regardless of financial constraints.

"We have a long-term vision through this process," Uludag said, "We intend to prepare tomorrow's workforce for emerging cybersecurity problems at an exponential scale."