As cyber threats continue to increase in complexity and scale, artificial intelligence (AI) is emerging as a critical tool for improving security, efficiency and accuracy within federal agencies, according to federal cybersecurity experts.
At GovExec’s Cybersecurity Futures Forum on November 20, cybersecurity leaders discussed how AI can help federal cybersecurity personnel conduct cyber defense operations more effectively and efficiently.
Paul Blahusch, chief information security officer at the Department of Labor, emphasized that AI will be essential for future cyber defenses because adversaries are already using the technology for AI-based attacks.
“Now we’re seeing adversaries start to use AI to attack, and so we’re going to have to fight the fire. We’re going to have to use AI to fight AI,” Blahusch said. “In the future, AI will definitely be necessary. »
For example, AI can help identify and locate anomalous data. Blahusch explained that AI can provide initial classification of data, helping to identify patterns.
“I could say to the AI: ‘Here are examples of important data – now go find other similar examples,’” he said. “We will need to analyze threat intelligence and correlate all the data to enable the automation that AI can drive for protection. »
James Saunders, chief information security officer at the Office of Personnel Management, also touted the use of AI for cyber defense operations. According to Saunders, AI provides federal agencies with the speed needed to address vulnerable solutions.
According to Saunders, AI can be incredibly useful in software development, acting as a “co-pilot” for developers by helping them break down tasks and streamline the process.
“Using AI and automation, we can act quickly by applying AI at every stage of the software quality chain, as well as within monitoring and package management teams, ensuring that we build in secure software from the start,” Saunders said.
He further explained that integrating AI and automation early in the secure software development process also helps check for weaknesses that developers are not able to detect in the code and base of real development.
