Artificial Intelligence

Carnegie Mellon Software Engineering Institute Forms AI Security Incident Response Team

• By Kate Lucariello
• 12/06/23

The Software Engineering Institute (SEI) at Carnegie Mellon University has created an Artificial Intelligence Security Incident Response Team (AISIRT) to analyze and respond to threats and security incidents involving the use of AI and machine learning (ML). The team will focus on dealing with threats from many different AI and ML systems, including commerce, lifestyle, and important infrastructure such as defense and national security, the SEI said. The team will also lead research into AI and ML incident analysis, response, and vulnerability mitigation.

The SEI noted that the rapid expansion of AI and ML platforms and software has presented serious safety risks from improper use or deliberate misuse. Prevention and mitigation of threats requires cooperation among academia, industry, and government, it said.

AISIRT will draw upon university cybersecurity, AI, and ML experts and work on furthering the recommendations made by the White House's Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, released in October 2023.

"AI and cybersecurity experts at the SEI are currently at work on AI- and ML-related vulnerabilities that, if left unaddressed, may be exploited by adversaries against national assets with potentially disastrous consequences," said SEI Director and CEO Paul Nielsen. "Our research in this rapidly emerging discipline reinforces the need for a coordination center in the AI ecosystem to help engender trust and to support advancing the safe and responsible development and adoption of AI."

This is not the SEI's first foray into cybersecurity, the institute said. Its CERT Coordination Center has been operating since 1988 to address vulnerabilities in computer systems. SEI also heads the National AI Engineering Initiative, and its experts are working on practices that support secure and human-centered AI.

Those who have experienced or are experiencing AI vulnerabilities or attacks may report them to AISIRT here.