Alert! Danger Lurks Behind AI Caricatures

LidahTekno, JAKARTA — The trend of creating caricatures and illustrations based on artificial intelligence (AI) is increasingly popular on social media. However, behind the attractive visual charm, there are risks that need to be considered. Cybersecurity company Kaspersky warns that this practice could lead to the exposure of personal information and allow the spread of highly personalized fraudulent messages.

In creating caricatures that are currently viral, users often share personal photos and ask the AI ​​system to create animations that depict life, profession or daily activities. These visuals are created based on data that AI knows about the user.

Adrian Hia, Kaspersky Asia Pacific Managing Director, stated that behind the popularity of this trend, there is the potential for disclosure of personal information. This could pave the way for large-scale, personalized fraud. According to him, this trend may seem like ordinary entertainment, but it actually provides information to cybercriminals.

According to Adrian, the request to create AI-based caricatures is not just about using visual filters. To make the results more accurate, users tend to provide broad access to personal information through instructions such as “make a caricature of me and my work based on everything you know about me.”

In addition to reference photos, users often include additional details such as company name, logo, job title, work location, daily routine, hobbies, and family information. Each of these pieces of data, when combined, can form a very detailed digital profile.

According to Adrian, the combination of images, text, as well as the context of personal and professional life can be used by cybercriminals to design more precise social engineering attacks. Information about place of work, position, or family members can make fraudulent messages seem more convincing and increase the likelihood of victims providing sensitive data or even transferring funds.

“Every time users in Asia Pacific provide details about themselves to AI just to see a clever illustration, they are handing over the blueprint for the perfect social engineering attack,” explains Adrian.

This risk is considered increasingly relevant in the Asia Pacific region. AI adoption rates in the region are high, with 78% of professionals using AI every week, surpassing the global average of 72%. However, uneven basic technical literacy makes some users vulnerable to phishing and social engineering-based manipulation.

On the other hand, interaction with the AI ​​platform does not stop with the final image output. Depending on the privacy policy of each service, original photos, text instructions, usage history, and technical data such as IP addresses, device types, and interaction patterns may be stored in the system. Some of this data may be used for service operations, performance improvements, or AI model training, so users’ digital footprints may potentially be retained longer than expected.

“In a region with leading AI adoption but lagging technical literacy, this digital portrait becomes a dangerous map,” stressed Adrian.

Along with these risks, Kaspersky experts recommend adopting digital hygiene when participating in the AI ​​caricature trend. To reduce risk, avoid entering identifiable data in fields, such as full name, job title, company, city, address, schedule, or routine, even if it appears to be just to personalize the image.

“Don’t upload photos that show logos, credentials, documents, vehicle license plates, screens, building facades, or any element that could help find you or associate you with an organization,” said Adrian.

He also advised against sharing information or images of minors, or disclosing family details that could be used to impersonate close contacts or devise emotional deception.

Then, review the platform’s privacy policy and permissions before using it, especially regarding content storage and use of data for training or service improvement. Supplement caution with active digital protection to reduce risks from malicious links, malicious downloads, and phishing techniques associated with this trend, while strengthening device security.