AI -managed attacks, become faster, sophisticated and difficult to detect by 2026, shows new Definea reports shows

San Francisco, April 03, 2025 (Globe Newswire) – Delinea, a leading supplier of decisions for providing human and machine identities with centralized authorization, has released its inauguration today Cybersecurity and landscape threats AI report. In a report based on exclusive Labs Delinea studies and the analysis of several third-party research and cyber innid in the real world, shows critical ideas about the rapidly developing landscape of threat. Since AI processes both attacking and defense strategies, the results emphasize the urgent need for organizations to adapt their security measures to counteract threats focused on AI.

The report reveals a sharp increase in attacks aimed at inhuman identities (NHI), such as services, APIs and machine identity. Delinea research shows that for each person a person is currently 46 NHI, a number that is expected to exceed 45 billion by 2025. Over 70% of NHI do not rotate in the recommended time frames, the average cycle of 627 days, which is much higher than the best security practices. In addition, 97% of organizations are subjected to their NGOs, which significantly increases the risk of unauthorized access.

“One of the biggest problems identified in the report is increasing targeting for inhuman identities,” John Kun said, SVP management of products in Delinea. “For organizations, this shift means that they face mass and often ignore the safety break. With the number of these machine identities, which are expected to grow exponentially in the coming years, as businesses continue to accept the AI ​​quickly, the lack of proper trusted management and the irradiation of these identities to the third parties creates serious vulnerability that cyber -inadants can operate to be used.

Redemption revival

The report also shows an alarming trend in redemption attacks, which in 2024 became more sophisticated with the increase in double extortion tactics. Cybercriminals not only encrypt, but also sensitive data on Efiltrate, threatening to release them publicly if the ransom is not paid. Five redemption groups – Ransomhub, Lockbit, Play, Akira and Hunters – were responsible for more than 36% of all the incidents of Redemption software controlled by Delinea in 2024, more than 5700 attacks. Looking ahead, the report provides a surge in the AI-controlled phishing-attacks, cybercriminals use AI to craft for an increasingly convincing phishing e-mail, which complicates the distinguishing legal communications from harmful.

“Increasing the sophistication of redemption and the increasing prevalence of AI-driven attacks are undeniable trends in today’s cybersecurity landscape,” said Gal Diskin, Vice President of Threat and Studies in Delinea. “Our studies show that cybercriminals are increasingly using AI and powerful tools for redemption software as RAAS) to start more purposeful and scaled attacks, especially around phishing and machine identities. dealing with a permanent remedy. “