The Rising Threat of AI-Driven Cybercrime

Recent research highlights a concerning trend: the surge in enterprise AI adoption has paved the way for a significant increase in AI-based tools tailored for illegal activities. These tools have become more affordable and accessible, allowing criminals of varying expertise to launch large-scale attacks with ease.

The Growing Danger of Deepfake Technology

Gartner analyst Dan Ayoub emphasized the escalating threat, stating, "The availability of high-quality GenAI applications now enables the creation of photo-realistic video content that can deceive or mislead audiences. Given the minimal barriers to using these sophisticated tools, a structured approach to detecting GenAI deepfake content has become essential."

It also reveal that the proliferation of deepfake technology and advanced AI tools in the cybercrime underground has opened doors for mass exploitation, even for cybercriminals lacking technical skills.

The Resurgence of Malicious AI Tools

The report also notes the return of defunct criminal Large Language Model (LLM) services like WormGPT and DarkBERT, now upgraded with new functionalities. These are being marketed alongside new offerings such as DarkGemini and TorGPT, which boast multimodal capabilities, including image generation.

However, the report cautions that many of the ChatGPT-lookalike services on the cybercrime underground are merely "jailbreak-as-a-service" front-ends, designed to trick commercial LLMs into providing unfiltered responses to malicious queries. Interestingly, cybercriminals have been slow to adopt these generative AI tools, likely because existing tactics and techniques are still effective without the need for new technologies.

The Need for Innovation in Cybersecurity

With the growing sophistication and frequency of cyberattacks, traditional cybersecurity measures are no longer sufficient. Businesses must proactively test their systems and networks to identify and address vulnerabilities before threat actors can exploit them. Additionally, individuals should exercise caution with unsolicited communications and verify the authenticity of online interactions.

Detecting and countering these AI-driven methods is crucial for better managing enterprise attack surface risk and reducing overall online risks for consumers. In a recent Trend Micro survey, 71% of respondents viewed deepfakes negatively, identifying fraud as one of their primary uses.