As AI technologies continue on a rapid-growth trajectory, some cybersecurity experts are worrying about how emerging versions of AI tech may harm the cyber-resilient architectures that many organizations are beginning to build.
During a webinar titled “Cyber Resilience for Modern, Agile, and Trusted Agency Environments” hosted by ATARC on May 25, Emily Odom, cyber assistant to the Special Agent in Charge at the Washington Field Office in the FBI, discussed the ease of use that AI technology offers to cybercriminals to carry out their illegal acts.
“The way I see ChatGPT as part of the cybercriminal ecosystem is it’s just going to make it easier for them,” she said. “Cybercriminals are already developing an ecosystem that allows sort of plug-and-play modules to do certain cyber-attacks. You no longer have to be extremely technical. You can buy these services to do these things. And I think ChatGPT is going to add to that,” said Odom.
Along with those worries, others see the need for cybersecurity officials and experts to step up and protect AI technology in order to ward off some of its possible downsides.
At the same webinar, Trent Frazier, deputy assistant director of the Stakeholder Engagement Division at the Cybersecurity and Infrastructure Security Agency (CISA), discussed the importance of protecting AI technology in order to curb its malicious use.
“How do you defend AI itself,” he asked. “There is a lot of interesting discussion about the fact that new AI capabilities or machine learning capabilities at a lower threshold are being employed in business and industry to automate any number of operational processes.”
“Much of that AI is not well defended, and I encourage folks to take a look at that,” said Frazier.
Ultimately, cyber defenders and experts can only do so much on their own, and may require more help from the government, said Felipe Fernandez, chief technology officer at Fortinet Federal.
“We need some governance and AI development and some control and regulations around it,” he said. “And move from there very smartly, quickly, of course, making sure that we’re not kind of just putting the brakes on AI development in the United States, but monitoring it, making sure that we’re certainly leveraging it to enhance our capabilities.”