Unsurprisingly, criminals have been exploiting open-source generative AI programs for various malicious activities, including creating malware and conducting phishing attacks, as stated by the FBI.
In a recent call with journalists, the FBI highlighted how generative AI programs, highly popular in the tech industry, are also fueling cybercrime. Criminals are using these AI programs to refine and propagate scams, and even terrorists are consulting the technology to develop more powerful chemical attacks.
A senior FBI official stated that as AI models become more widely adopted and accessible, these cybercriminal trends are expected to increase.
Although the FBI did not disclose the specific AI models used by criminals, it was revealed that hackers prefer free, customizable open-source models and pay for private hacker-developed AI programs circulating in the cybercriminal underworld.
Seasoned cybercriminals are exploiting AI technology to create new malware attacks and improve their delivery methods. For example, they use AI-generated websites as phishing pages to distribute malicious code secretly. The technology also helps hackers develop polymorphic malware that can bypass antivirus software.
Last month, the FBI issued a warning about scammers using AI image generators to create sexually themed deepfakes to extort money from victims. The extent of these AI-powered schemes remains unclear, but the majority of cases reported to the FBI involve criminal actors utilizing AI models to enhance traditional frauds, including scams targeting loved ones and the elderly through AI voice-cloning technology in phone calls.
In response, the FBI has engaged in constructive discussions with AI companies to address the issue. One proposed solution is using a "watermarking" system to identify AI-generated content and images more easily.
The senior official emphasized that the FBI considers this AI threat a national priority, as it affects all programs within the agency and is a recent development in the cybercrime landscape.
