The increasing integration of artificial intelligence (AI) across various industries has concurrently resulted in a rise in the number of vulnerabilities identified within AI services, as detailed in Kaspersky’s Quarterly Exploits and Vulnerabilities Report.
Kaspersky’s Vulnerability Management data indicates that during the second quarter of 2026, the reported vulnerabilities in AI and large language model (LLM) services surged tenfold since the fourth quarter of 2025, totaling 935 vulnerabilities.
Alexander Kolesnikov, a Malware Expert at Kaspersky, stated, “While tools, plugins, and AI technologies are instrumental in automating tasks efficiently, their rapid deployment within organizations introduces new cybersecurity challenges. Notable vulnerability types include weaknesses in access control, authentication processes, and injection handling. To mitigate risks, organizations must ensure real-time oversight of their infrastructure and access controls, rather than relying solely on traditional patch management, enabling them to identify and address threats promptly.”
Submit Your Press Release
Get your company news, announcements, launches, appointments and events in front of a wider audience.
In the second quarter of 2026, the count of critical vulnerabilities reached 119, nearly five times the figure reported at the end of 2025. The complete report can be accessed on Securelist.
Analysis of prevalent vulnerability categories reveals that significant problems in AI-related software stem from inadequate access control mechanisms for vital system components, flawed implementation of authentication and authorization processes, and injection vulnerabilities.
Furthermore, the recent Kaspersky SMB report highlights that over 33,300 attacks targeting small and medium-sized enterprises were detected in the first four months of 2026, where malicious software was disguised as popular AI services. This figure represents a nearly fivefold increase compared to the same timeframe in 2025, underscoring the correlation between the growing use of AI services and the rise in cyberattack attempts against them.
