Sign up to save your podcasts
Or
Share S2 Ep. 5 AI, Machine Learning & Strengthening Cybersecurity
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
S2 Ep. 5 AI, Machine Learning & Strengthening Cybersecurity
Interest in machine language and artificial intelligence has been growing and growing since around 2015. It was a perfect storm where storage prices decreased, and virtualization became standard. Like most maturing industries, challenges appeared. It was found that the conclusions provided by artificial intelligence were dependent on the quality of the data it collected. This finding became so common that in 2020 the federal government responded with Executive Order 13960 promoting the use of trustworthy intelligence in the federal government.
This is an interview with a focus on applying that guidance across many federal areas, including cybersecurity.
Technology leaders from the National Science Foundation have formulated the National Artificial Intelligence Research Resource Task Force (NAIRR) https://www.nsf.gov/cise/national-ai.jsp to give guidance on using data more effectively. Recommendations include the categories of security privacy, civil rights, and sustaining the resource. The goal was to help get control of the unwieldy and expanding IT environments.
Wayne LeRiche, Palo Alto Networks, gives an example of applied AI with some of the security concerns relating to Domain Name Servers (DNS). Attacking a DNS server is a classic approach malicious actors have used for years. The traditional method of defense is rule-based. If “x” occurred, then react with “Y.”
Attackers can ratchet up the attack with something unknown that the rules-based system can’t handle it. AI can handle drastic increases in speed that a rules-based system can’t. Further, artificial intelligence can understand unknown methods of denying service to the DNS.
Rather than focusing on one aspect of prevention, experts on the panel suggest that Rule-Based and Machine-Learning DNS security services be used in parallel.
Finally, Tony Walker, NetScout, presents a scenario where one part of the team is proficient in network management while another part of the team has a sophisticated knowledge of data management. The solution that is provided by AI must be tempered with careful consideration of data sources and the ability to use that knowledge in an environment full of people with different skill sets.
View all episodes
By FedInsider
5
55 ratings
Interest in machine language and artificial intelligence has been growing and growing since around 2015. It was a perfect storm where storage prices decreased, and virtualization became standard. Like most maturing industries, challenges appeared. It was found that the conclusions provided by artificial intelligence were dependent on the quality of the data it collected. This finding became so common that in 2020 the federal government responded with Executive Order 13960 promoting the use of trustworthy intelligence in the federal government.
This is an interview with a focus on applying that guidance across many federal areas, including cybersecurity.
Technology leaders from the National Science Foundation have formulated the National Artificial Intelligence Research Resource Task Force (NAIRR) https://www.nsf.gov/cise/national-ai.jsp to give guidance on using data more effectively. Recommendations include the categories of security privacy, civil rights, and sustaining the resource. The goal was to help get control of the unwieldy and expanding IT environments.
Wayne LeRiche, Palo Alto Networks, gives an example of applied AI with some of the security concerns relating to Domain Name Servers (DNS). Attacking a DNS server is a classic approach malicious actors have used for years. The traditional method of defense is rule-based. If “x” occurred, then react with “Y.”
Attackers can ratchet up the attack with something unknown that the rules-based system can’t handle it. AI can handle drastic increases in speed that a rules-based system can’t. Further, artificial intelligence can understand unknown methods of denying service to the DNS.
Rather than focusing on one aspect of prevention, experts on the panel suggest that Rule-Based and Machine-Learning DNS security services be used in parallel.
Finally, Tony Walker, NetScout, presents a scenario where one part of the team is proficient in network management while another part of the team has a sophisticated knowledge of data management. The solution that is provided by AI must be tempered with careful consideration of data sources and the ability to use that knowledge in an environment full of people with different skill sets.