AI-powered attacks outpace cloud defenses

Cloud security faces growing challenges as attackers increasingly leverage automation and AI to exploit vulnerabilities, dramatically reducing the time needed to steal data or compromise systems, according to Computer Weekly.

Sergej Epp, CISO at Sysdig, warned that bots are now the top threat in the cloud, with "hackbots" becoming faster and more sophisticated, sometimes exfiltrating data in under five minutes. Epp stressed that while defenders know what needs to be done, acting fast enough remains a problem, especially for under-resourced firms. He emphasized that the focus should shift from AI models to securing the underlying infrastructure, noting that existing methods like firewalls and static scanning are ineffective against AI-related threats such as prompt injection. Epp advocated for constant asset monitoring, runtime security agents in every container, and intelligent filtering of telemetry data due to the cloud’s fleeting nature. While Sysdig promotes automation through its Sage AI analyst, Epp cautioned that defensive errors carry a far higher cost than attacker failures.