Does Genai shortcuts make attackers lazy and make them easier to capture?

“In short, the AI-driven offense is real, but still a bit clumsy, and the transparency of model providers turns clumsy detectives into detection advantages,” Roberts said. “The security team should impose similar reports on suppliers and connect these metrics to the next supplier.” [genAI-fueled attack] Appear. ”

Attacker’s Tactics

The OpenAI report, released in June, details the company’s various defenses deployed by fraudsters. For example, one of them involves a forged assignment application.

“We identified CHATGPT accounts related to employment activities that appear to have multiple suspiciously deceptive. These threat actors use OpenAI’s model to develop materials that support fraudulent attempts to apply for it, software engineering and other remote work in the world,” the report said. “While we cannot determine the location or nationality of threat actors, their behavior is consistent with the public activities of IT-related workers’ programs with North Korea. Workers’ programs (DPRKs) related to North Korea. Some participants related to these latest campaigns may have been used by takeovers as contractors, including DPRK at the core core, to perform application tasks and operate hardware, including scopes of use within us, including our scope.”