The rise of generated AI and large language models has completely changed the cybersecurity situation. Empower attackers with easy-to-use tools This creates realistic video and sound deep effects, personalized phishing activities, and malware and malicious code.
This also opens the door for AI to defend. As agent AI becomes more deeply embedded into the enterprise Finance and other fields and LegalCybersecurity AI Agents are also rising, becoming a key asset for detection, analysis and alerting.
“Discover, containment, investigation and response are a huge challenge,” said Brian Murphy, CEO of Reliaquest, a cybersecurity technology company. “AI allows us to remove a lot of noise, namely, first-level or second-level jobs that are often completely unrelated to what might threaten the organization.”
Taking tools into the hands of humans, these tools can automate otherwise causing tasks or time-consuming tasks, freeing them to do more important work, which is often the tendency of Agesic AI.
exist News shared with Amazon employees in June“We strongly believe that AI agents will change the way we work and live,” said CEO Andy Jassy. He added that he sees the future of “billions of these agents, in every company and in every imaginable field,” “help workers” focus on rote work, while thinking strategically while making our work even more exciting, “make us more exciting and fun.”
Murphy has a similar view on cybersecurity, seeing an industry of workers flooded with their jobs, they probably shouldn’t spend time, causing more burnout and exacerbating existing lack of available talent.
He also saw AI’s way of attacking companies. “Those phishing emails that used to look almost ridiculous, spelling and font-errors,” he said. “AI can bring normal bad actors and make them better, so the trick is that if you’re on the defensive side, they have to use AI because of what AI can do.”
Reliaquest recently released the so-called Greens Agent Teammate, a role-based AI agent that can be used to perform tasks that detection engineers or threat intelligence researchers implement in the security operations team.
“Think of it as a role of unity with humans, and humans are driving human AI, so humans know what to do,” Murphy said.
Murphy gives an example of any security team for a global company: international executive travel. Every time you connect your laptop or phone to a network in China, the security operation team is reminded that the security team must verify executives are abroad and use their devices firmly. Using proxy AI teammates, the security staff can automate the task and even set up a series of similar processes for board meetings, on-site or other large team gatherings.
“There are hundreds of things,” he said.
Justin Dellaportas, chief information and security officer at communications technology company Syniverse, said that while AI agents have been able to automate some of those basic cybersecurity tasks like combining through logs, it’s also starting to be able to automate actions, like quarantining flagged emails and removing them from inboxes, or restricting access by a compiled account across a variety of logins.
“Criminals are using (AI) to effectively detect vulnerabilities and large-scale organizations, all of which lead to higher success rates, faster access and accessing organizations faster than we see,” he said. “Cyber defenses actually need to rely on this technology more than ever before to keep this evolving threat landscape and the pace of cybercriminals.”
Dellaportas said that while each company has a unique risk profile and tolerance when it comes to deploying different types of cybersecurity tools, he believes that the adoption of proxy AI in cybersecurity is the stage of “crawling, walking, running methods.”
“You roll it out and then take action, but then iterate through the actions you’ve taken before,” he said. “I’m back to a trust, but to verify, and then as we have confidence in its effectiveness, we’ll continue to solve different problems.”
What AI robots mean for cybersecurity workers
Dellaportas said AI agents can take over certain tasks of human cybersecurity professionals in the future, but he still believes the technology is an enhancement rather than a replacement for workers.
Murphy agreed, saying he did not see proxy AI replacing actual cybersecurity workers, but helped complete tasks where automation is a better choice, while also addressing the skills gaps many organizations struggle with when they fill their cybersecurity roles.
“It may be a lack of trained and skilled cybersecurity professionals, but there are no shortage of people who want to be trained and skilled in cybersecurity,” he said. “The reason why knowledge transfer takes a long time to network is that when you get entry-level jobs, it’s the equivalent of working on the help desk.”
Murphy said he learned that it still takes a lot of education when deploying proxy AI in any part of the business and was concerned about how AI decisions are made.
Dellaportas said the fact that it helps all types of business lines use proxy AI, so it is not new to discuss how these AI tools can help achieve their goals.
Artificial intelligence agents are catching up with internal companies. A May 2025 poll of 147 CIOs and leaders in Gartner operations found that 24% have deployed several AI agents, with more than 50% working in areas such as IT, HR and accounting, while only 23% of the features faced by external clients.
In the field of cybersecurity, companies trying Agentic AI find it “moderately beneficial”, but there are still some problems with the ability of these tools to extend beyond simple tasks, said Avivah Litan, a distinguished vice president analyst at Gartner’s AI strategy team.
“Security has always been a fruit use case for AI,” Litan said. “You first see fraud detection in AI, so we’ll get 100% of digital security assistance in the future and free up employees to do new attacks; the key is to make sure they stay up late so they see all this innovation so they see the entire attack surface.”
Murphy believes that business adoption and proxy AI for cybersecurity may develop faster than financial or legal legality.
“They absolutely understand that AI is being dealt with, and the only way to defend is to use it in their own defense,” he said.
