Cybersecurity professionals say generative AI can be exploited in cyberattacks — but it can also be a powerful defense

• Generative AI is revolutionizing cybersecurity in connected networks.
• Companies that don't adopt AI risk falling behind in cybersecurity, IT experts warn.
• This article is part of "5G and Connectivity Playbook," a series exploring some of our time's most important tech innovations.

Generative AI has become a double-edged sword for the security of connected networks.

On one hand, generative AI can speed up cybersecurity problems, making it easier and cheaper for bad actors to conduct identity attacks. For example, it can be used to design sophisticated phishing campaigns by generating audio, images, or videos to create fake identities.

While hackers exploit AI technologies, companies are adopting AI as a defense tool. As the number of connected devices on 5G networks increases, generative AI in security operations will become more crucial, David Cooper, a global lead for Accenture Security, said.

"What used to take a human being two hours, now there's an AI solution to do it in much faster time," Cooper said. "We're able to respond faster, we're able to catch up to the bad guys faster with interesting transformational technology. We're able to move so much faster and do more with limited resources."

Business Insider spoke with several cybersecurity professionals who attended the annual RSA conference this week in San Francisco about how generative AI is increasingly used in cybersecurity for 5G devices and networks. They said that while 5G networks have higher security standards, attacks could still come from identity breaches of accounts or devices connected to those networks.

"If you can stay ahead of the risks you perceive when you implement 5G solutions, you're going to be in good shape," said Shaun McAlmont, the president and CEO of the cybersecurity-awareness-training company Ninjio.

Accelerating cybersecurity

Generative AI can speed up more routine security processes and free up resources, helping cybersecurity professionals increase productivity and make decisions faster. Things like the design and architecture of a 5G application, generating code samples, conducting tests, and speeding up authentication can be automated. It can also prevent attacks by helping to configure and manage networks.

"There's a distinct benefit for organizations where we may leverage AI for improving network quality, improving security, and network healing," Chris Novak, the senior director of cybersecurity consulting at Verizon Business, said. "There are opportunities for AI to help us analyze network traffic flows in real time, recognize where there is an issue, and ensure little to no impact to network users."

AI can also help 5G users monitor their security and ensure that the services they connect to have the appropriate certifications and authorizations for transferring data.

Large language models can help mitigate security threats by identifying suspicious patterns in traffic and unusual attempts to enter the network. Because AI can process an enormous amount of data about malicious threats, it can help isolate them before they spread further. Additionally, it can triage malware and other threats, making sense of complex attack-path data and automating threat responses.

Edge-protection tools powered by AI can be a "first layer of defense" that can deal with mass data, David Aviv, the chief technology officer at Radware, said.

"Competitors that didn't adopt those machine-learning technologies have gotten left behind," said Steve Wilson, the chief product officer at Exabeam, a network security company.

5G to strengthen AI

5G can also boost the capabilities of AI applications themselves. With 5G, AI can do more computation and sensor analysis on the edge, meaning the computation is closer to the data source.

"The combination of AI and 5G will make it inevitable that we will be using machines for tasks that are currently performed by humans," said Ev Kontsevoy, the cofounder and CEO of Teleport, an identity access and management software company.

It's becoming increasingly important for companies to apply AI to their cybersecurity practices because humans won't be able to keep up, Rohit Ghai, the CEO of RSA, said.

"Any 5G network is very very dynamic," Ghai said. "It's changing all the time. Human approaches to securing that are not able to keep up with the dynamic nature of the network."

Now that mobile and Internet of Things devices have become more ubiquitous, they're using even more data, which generative AI can help filter and secure.

"We're going to be required to collect an order of magnitude more data from more devices and provide really quick analysis on that," Wilson said.