Leading researchers in AI recently came together at Embry-Riddle Aeronautical University to discuss issues around cybersecurity, trust, verification, and practical applications across fields.

The Fourth Lawrence H. Harper Memorial Workshop on Cybersecurity and AI drew more than 50 participants and featured 17 presentations from speakers around the world.

Dr. Jon C. Haass, a professor of cyber intelligence and security and organizer of the Aug. 14-15 event at the Prescott Campus, said the workshop reflects Embry-Riddle’s growing emphasis on AI as both a research area and a skill set students will need in the workforce.

“This workshop brings practitioners of AI for practical real-world problem-solving in a variety of sectors,” he said. “It is important to Embry-Riddle as we continue to integrate AI concepts into our curriculum as another tool and set of skills our students will need in their careers.”

The workshop is named for the late Professor Lawrence H. Harper, who passed away in October 2023. Harper worked at UC Riverside in California and collaborated with Embry-Riddle adjunct professor Joe Barr on an outstanding problem in graph theory. He also worked in information transmission and encryption and spent years conducting research at NASA’s Jet Propulsion Laboratory. He served as the keynote speaker for the first AI workshop in 2022.

Dr. Dick Bulterman, a leading AI researcher from the Netherlands, served as this year’s keynote speaker.

“There’s a lot of technology in AI, and usually it works surprisingly well. It works uncomfortably well. It gives a feeling of the sorcerer’s apprentice,” Bulterman said. “We’re in a phase of our work with AI that’s a lot like the infatuation phase in a relationship. But where is it taking us, and what is it going to do for us?”

The program also included a presentation by Frans Bulterman, the son of keynote speaker Dick Bulterman and a researcher at Vrije Universiteit Amsterdam. He spoke about the cognitive aspects of using AI and the growing tendency to trust AI systems over firsthand observation.

“The results of these systems give, they’re so good and they’re basically to a point where now a lot of people trust them over their own instruments,” he said. He added that people can end up “relying on a chatbot to tell them if the world they see around them is actually what they think.”

Another session examined how to turn competing signals into a trustworthy decision. David Ostrowski, a researcher at Northwestern University in Evanston, Illinois, said the challenge is not simply collecting data but making sense of it.

“Ultimately, you need an answer, right?” Ostrowski said. “The consensus mechanism is going to present the largest challenge.”

Haass said the larger challenge facing AI is building systems that are useful but also explainable and accountable.

“The analogy of shining light from all directions means no shadows,” he said. “We can see problems and find solutions that interdisciplinary approaches provide.”

The workshop concluded with a look ahead as organizers discussed future collaborations in AI and cybersecurity research.

“This type of workshop allows students to present their work, speak to industry and government professionals and learn what future careers may be,” Haass added. “They can learn from the current leaders how they navigated and rose in their own careers with panels of professionals in different roles.