A group of cybersecurity experts has issued serious warnings about the wave of AI-generated caricatures, which rely on analyzing what a chatbot knows about a user to transform their identity into an artistic image.
The experts explained that this trend, which seems entertaining on the surface, could in reality be a “goldmine of information” for fraudsters; uploading personal images coupled with job details or company logos gives digital criminals a precise visual and informational representation of an identity, making it easier for them to impersonate individuals or create highly realistic “deepfake” videos.
It was pointed out that intelligent systems do not just produce the drawing, but also process the image to extract sensitive data including emotions, the surrounding environment, and geographical location, with the potential to store it for unknown periods to train future models.
–
–
–
User engagement in these steps was described as “doing the fraudsters’ work for them,” with warnings that the way these trends are promoted raises alarm bells, sometimes appearing as if they are “deliberately designed by hackers” to compile massive databases on employees and companies.
Expert Advice to Avoid Risks
To avoid these risks while continuing to keep up with modern technologies, experts advise carefully reviewing privacy settings and enabling the “do not train on my content” option available on platforms such as OpenAI.
It is also strongly emphasized to reduce details in uploaded images by “tightly cropping the photo” and hiding any indications of uniforms or location, while avoiding sharing precise job titles in operation commands. It was confirmed that a single high-resolution image in the “wrong hands” is enough to open wide doors to complex fraud operations that go beyond just traditional account theft.
