AI agents have developed in the forces of protection of cybersecurity companies

The growth of generative models of II and large language models dramatically shifted the cybersecurity landscape, Expanding attackers with easy -to -use tools This can create realistic videos and voice deep features, personalized phishing companies, malware and malware.

It also opened the door to the defense. As the Agentic AI becomes deeper into the enterprise in the Areas such as finances and legalAI cybersecurity agents are also growing, becoming a key asset for detection, analysis and announcements.

“This is a big problem for identifying, maintaining, investigating and responding in large companies,” said Brian Murphy, CEO of Cybersecurity Reliaquest technology. “AI allows us to remove a lot of this noise, this level of the first or two levels that are often not relevant to what the organization may threaten,” said Murphy.

Putting the tool into the hands of human workers, which can automate the main tasks or time -consuming times, freeing them to do more important work, often became for Agentic AI.

In A The message shared with Amazon staff in JuneThe CEO Andy Yassi said: “We have the belief that AI agents change the way we all work and live,” adding that he sees the future with “billions of these agents, in each company and in each imaginary field,” helping workers “will focus less at work, and more.

Murphy shares a similar opinion on cybersecurity, where he sees a branch of workers who are littered with work, most likely not waste time, causing more burnout and increasing the existing problem of absence of available talents.

He also saw AI’s attack on the company. “These phishing leaves, they looked almost funny with the wrong writings and fonts wrong,” he said. “AI can take the average bad actor and make them the best, and therefore the trick is when you are in the defensive side, they should use the II because of the reality of what AI can do.”

Recently, Reliaquest has released what he calls Greymatter’s teammates, autonomous, role -playing agents that can be used to accept the tasks that the exploration engineers or researchers otherwise performed on security team.

“Think about it as this person who unites the person, and the person pushes this agent II, so the person knows what to do,” said Murphy, adding that it is like a “teammate who accepts the response analyst at the incident and multiplies their opportunities.”

Murphy gave an example that is a frequent phenomenon for any security team in a global campaign: international leadership. Whenever the laptop or cell phone is connected to the network in, say, China, the security team will be warned, and the security team will have to make sure that the executive power is abroad and reliably uses their device every day. With the help of Agentic AI teammates, a security person can automate this task or even create a number of similar processes for council meetings, outside of sites or other major team meetings.

“There are hundreds of such things,” he said.

Justin Deloportas, Chief Information and Security Service Communications Technology Company Syniverse, said while AI agents were able to automate some of these basic cybersecurity tasks, such as combing magazines, he also begins to automate actions like, for example, Coantic Electronic Leaves and An account consisting of different inputs.

“(AI) criminals use to effectively search for vulnerabilities and feats in the organization, and all this leads to the fact that they have a higher level of success, having gained initial access and go aside to the organization faster than we saw,” he said. “The cyber -defender really needs to lean against this technology, so that the landscape of the threat and the tempo of cyber -ears develops.”

Dellaportas said, while each company has a unique risk profile and tolerance when it comes to deploying different types of cybersecurity instruments, it views the acceptance of Agentic AI in cybersecurity stages “crawling, walking, launch methodology.”

“You unfold it, and it will think, and then take action, but then it must overcome the actions it was taken earlier,” he said. “I return to a peculiar trust, but I check, and then, if we are confident in its efficiency, we will go to different problems.”

What means AI boots for cybersecurity workers

While Dellaportas said AI agents may take some tasks from cybersecurity professionals in the future, he still sees this technology as an increase to make workers more effective rather than a replacement.

Murphy agrees and said he does not see Agentic AI, which takes the place of actual cybersecurity workers, but helping with the tasks when automation is the best option, and decides the gap in the skills with which many organizations are fighting when filling in cybersecurity roles.

“There may be a lack of trained and qualified cybersecurity professionals, but there are not enough people who would like to train and qualify for cybersecurity,” he said. “The reason that the transfer of knowledge takes so much time to cyber is that when you get an entry -level job, it’s equivalent to work on reference.”

Murphy said he realizes that he still needs a lot of education when it comes to deploying Agentic AI in any part of the business, as well as concern how decisions are made by AI.

Dellaportas said what helped is that Agentic AI is used by all types of business, so discussing how these AI instruments can help achieve goals that are not new.

AI agents catch internal companies. The poll in May 2025 at 147 CIOS and IT -Function Gartner revealed that 24% had already launched several agents of II, more than 50% of those AI agents as HR and accounting, compared to only 23% of customer features.

Avivah Litan, an excellent vice -president of AI Gartner Strategy Group, said in the Cybersecurity Space that experimenting with Agentic AI thinks “moderately” but remain some questions about the ability of these tools to scale outside the simpler tasks.

“Security has always been the use of fruit for AI,” Litan said. “You first saw how the II appeared with the detection of fraud, so 100% we will have digital safety assistance in the future, performing work and releasing employees to take new attacks; the key will remain with all these innovations so that they can see the entire surface of the attack.”

Murphy believes that the corporate adoption and evolution of Agen AI in cybersecurity can occur even faster than in finance or legal.

“They absolutely understand that the II is used against them, and the only way to protect is that it use it in your defense,” he said.

Source link