Image: Federal News NetworkFighting AI with AI requires enduring, new approaches
• Federal and industry experts are advocating for the use of continuous monitoring, evaluation, and "red teaming" to secure AI models against evolving threats. • Kevin Walsh, IT and Cybersecurity Director for the Government Accountability Office, warns that the rapid speed of AI-driven attacks is making it increasingly difficult to maintain human oversight. • To combat these risks, experts recommend implementing a universal data fabric for faster incident detection and a zero trust architecture to protect data pushed to the edge.
federalnewsnetwork.com
