As the promise of AI grows, so does the danger that criminals will find ways to take advantage of AI deployments, turning innovation and progress into opportunities for data breaches, theft, election manipulation and other forms of expensive, dangerous and embarrassing malfeasance. — Photo by Philipp Katzenberger on Unsplash
Two of the biggest issues in government technology are headed for a collision, one that requires the full attention of public-sector leaders, according to speakers at the State of GovTech 2024 conference in Kansas City, Mo.
Few gov tech topics command the focus of IT professionals as artificial intelligence and cybersecurity, trends reinforced by two recent survey reports.
But as the promise of AI grows, so does the danger that criminals will find ways to take advantage of AI deployments, turning innovation and progress into opportunities for data breaches, theft, election manipulation and other forms of expensive, dangerous and embarrassing malfeasance.
That was among the main messages this week at the conference, which attracted more than 160 people from the public sector, vendor and investment worlds.
"The pace of change is fast, but the danger of that is making sure it is used responsibly," said Tom Lynch, CIO of Cook County, Ill.
All types of public-sector tasks, from law enforcement to permitting to community engagement, are gaining the technological support of generative AI, with more such advancements certain to follow in the next few years.
"AI is something that everyone has their hands on," said Gail Roper, CIO for Montgomery County, Md.
That includes criminals.
As Lynch noted during a panel presentation, the "bad guys" are adopting the tech more quickly than public-sector tech departments, giving criminals an "inherent advantage".
One doesn't need to be an expert hacker to get access to cutting-edge AI tools, he said. He provided an example of the growing threat: While internal phishing tests in Cook County have generally shown security improvements and increasing awareness of threats among employees, the "last few months" have shown that criminals are making significant improvements with fooling people via phishing.
"It's a free-for-all world; the wild west," is how he described it before underscoring the stakes involved. "We have the most sensitive data of any industry in the world."
That doesn't mean the public sector is helpless when it comes to securing its AI efforts, of course.
Data management, including training models for AI, stands as a vital security task, according to conference speakers. That essentially means keeping tabs on how people are using data within an organisation, Roper said.
Working with and paying attention to vendors also are key, according to Lynch.
"Vendors are incredibly important, because (they) have access to our data or hold pieces of our data," he said. – Government Technology/Tribune News Service
