Cybersecurity expert Harriet Pearson talks implications of widespread artificial intelligence use

–>

Shepherd University President Mary J.C. Hendrix, left, stands with Axia Advisory LLC founder and CEO Harriet Pearson after the Sept. 24 lecture in the Byrd Center auditorium. Tabitha Johnston

SHEPHERDSTOWN — The auditorium was packed to the brim on Tuesday night in the Robert C. Byrd Center for Congressional History & Education, as community members and Shepherd University students gathered together to learn about the implications of the widespread use of artificial intelligence (AI).

Axia Advisory LLC founder and CEO Harriet Pearson, of Shepherdstown, spent a little over an hour discussing “Science Fiction or Fact? Artificial Intelligence Risk and What To Do About It.”

“AI [has] a very broad definition. It includes basically anything where computers act on tasks or perform that we would all normally think would require human intelligence,” Pearson, who currently works as a cybersecurity and data governance consultant, said. “The history of AI goes back to the 1950s.”

Informed by her experience as an adjunct lecturer for the University of California, Los Angeles School of Law and Georgetown University, Pearson simplified the subject of AI as much as possible, to ensure she achieved her goal — that her audience would retain a good portion of the lecture content.

“Science fiction has taught us what can go wrong,” Pearson said, mentioning one of her favorite films, “2001: A Space Odyssey,” particularly came to mind in regard to this subject. “AI is about data — you’re gathering information and then you’re taking software that really is doing mathematical calculations according to rules. That’s an algorithm. And the algorithm is operating on the data.

Community members and Shepherd University students mingle after the President’s Lecture in the Byrd Center on Sept. 24. Tabitha Johnston

“There are literally thousands of risks,” Pearson said. “What goes into the box where the AI is operating? What comes out — what does the output do or what is it capable of doing? That’s really at a system level. And then, broader than that are the system-wide risks — how do you govern that?”

Pearson noted that every industry will likely be both positively and negatively effected by the recent growth in AI software. Two detail-oriented industries — the medical and legal fields — she said will likely be the slowest adopters of the technology, due to its potential inaccuracies.

“Is data free from bias?” Pearson asked, mentioning this is particularly problematic, as AI is increasingly used to make personnel and other important work-related decisions. “What makes that data free of any of the inherit bias that we have as humans, whether it be gender or race or something else?”

In her own use of AI, Pearson said she has been made aware of another issue with AI’s accuracy — if the data is not available, the software will make information up.

“And for every error or gap in information, there can also be malfeasance — was the data poisoned? I’ve worked a lot on cybersecurity specifically, and this is one of those areas that’s a nightmare that really hasn’t come to pass yet. But it’s a nightmare we worry about,” Pearson said. “What happens if bad actors deliberately poison information or inject incorrect information?”

Axia Advisory LLC founder and CEO Harriet Pearson speaks in the Robert C. Byrd Center for Congressional History & Education on Sept. 24. Tabitha Johnston

For the public, Pearson summed up how best to interact with AI in three words.

“Be very careful,” Pearson said, mentioning this caution is especially important to follow this year, as AI-generated images and information spread misinformation about political candidates online.