U.S.: Experts bemoan lack of technically trained cybersecurity personnel
Washington, 6/19/2022
In early May 2022, the U.S. Senate Armed Services Committee’s Subcommittee on Cyber Affairs held a congressional hearing on artificial intelligence and machine learning in cyberspace. The committee discussed the topic with representatives from Google, Microsoft, and Georgetown University’s Center for Security and Emerging Technology.
In his introduction, the hearing’s chairman, Senator Joe Manchin (D-WV), emphasized the enormous importance of AI and machine learning to the United States Armed Forces. However, he said that the fact that there is a need for a backyard in this area is clearly due to the fact that there is a „shortage of technically trained cybersecurity personnel in both government and industry“ across the country.
Google cited three reasons why machine learning and AI are important to cybersecurity. According to Dr. Andrew Moore, director of Google Cloud Artificial Intelligence,da are first, the use of AI to defend against adversary attacks, second, and third, the organization of data and people. According to the expert, AI is capable of processing millions of attacks every second and responding to them. This exceeds the ability of any human in terms of processing data.
Turning to the human side of the equation, Moore emphasized that when „new attacks occur, humans ingeniously develop new methods and AI develops new methods, so you have to learn new patterns or detect entirely new types of attacks in real time.“
He turned to the topic of insider threats and emphasized the importance of AI in implementing zero trust, where AI can be used to detect human behavior patterns. Moore clarified that the quality of the solution depends on the scope as well as the quality of the data. and negative by an AI without data is „pretty worthless.“ He pointed out that isolated data is the nemesis of AI and that a full exchange of disparate data sets is needed to get a more complete picture.
Eric Horvitz , Microsoft’s chief scientific officer, also called the shortage of cybersecurity personnel problematic.Microsoft, he said, is committed to highlighting the company’s efforts to implement in line with President Biden’s executive order to improve the nation’s cybersecurity (EO 14028). In his opening remarks, he said, „The value of leveraging AI in cybersecurity applications is becoming increasingly clear. Among many other capabilities, AI technologies can enable automatic interpretation of signals generated by attacks, effectively prioritize threats, and provide adaptive responses to the speed and magnitude of adversary actions. The methods show great promise in quickly analyzing and correlating patterns across billions of data points to detect a variety of cyber threats in seconds.“
Horvitz emphasized that the cybersecurity workforce shortage is problematic, citing the Cybersecurity Workforce Study 2021, which puts the number of unfilled positions worldwide at 2.72 million. Even with operations teams working around the clock, there are far more alerts to handle than staff, leading to the very real risk that teams will be overwhelmed. AI, Horvitz said, „enables defenders to effectively scale their protection capabilities, orchestrate and automate time-consuming, repetitive and complicated response activities.“
The benefits of AI on the security side of the equation fall into four groups, according to Horvitz: Prevention, Detection, Investigation and Remediation, and Threat Intelligence. But AI-powered cyberattacks are also a reality. Criminal/national state attackers use basic automation, authentication-based attacks and AI-powered social engineering. His comments on „adversary AI“ made clear the need for continued investment in research and development to increase the robustness of systems. He was emphatic about the importance of red-team exercises.
Center for Security and Emerging Technologies: Focus on trustworthiness of AI systems.
Georgetown’s CSET was represented by Dr. Andrew Lohn, who is a senior fellow on CSET’s CyberAI project. He addressed three important areas of AI:
– AI promises to improve cyber defense.
– AI can improve offensive cyber operations.
– AI itself is vulnerable.
