A new AI-powered art trend has taken social media by storm, allowing users to transform their photos into stunning Ghibli-style illustrations. The technology, made widely accessible by OpenAI’s latest ChatGPT update, has captivated millions with its ability to mimic the signature aesthetic of Hayao Miyazaki’s animated worlds.
However, alongside the excitement, privacy experts are raising serious concerns.
Privacy warnings from experts
Proton, a data security platform, has issued a strong caution against the trend, emphasising the dangers of sharing personal images with Artificial Intelligence (AI) systems.
In a post on X, Proton warned, “Think this is a fun trend? Think again. While some people don’t mind posting selfies online, the rise of AI-generated Ghibli-style images means many are unknowingly providing OpenAI with photos of themselves and their families.”
Proton elaborated on these risks in a series of posts, highlighting the potential consequences of feeding personal images into AI. The platform noted that once a photo is uploaded, users effectively lose control over how it might be used.
“Aside from the risks of data breaches, once you share personal photos with AI, you lose control over their use, as they can be repurposed for training AI models. In some cases, these images could even be manipulated for harmful purposes, including harassment or defamation,” the post added.
“Many AI models, especially those used for image generation, depend on vast datasets. In some instances, photos of individuals may be incorporated into these datasets without explicit consent,” the platform stated.
Additionally, Proton warned that such data could be exploited for targeted advertising or even sold to third parties without users’ knowledge.
In an in-depth blog post, Proton explained how AI models like ChatGPT process vast amounts of data to improve their functionality.
“Massive datasets are essential for training AI systems. The more data they receive, the better they become at recognizing patterns and generating new content,” the post read.
“AI systems will only learn from the datasets they have access to. If those datasets contain biases, the AI will inevitably reflect and perpetuate those biases in its decision-making,” it added.
With the rapid rise of AI-generated content, privacy advocates continue to urge users to be cautious about sharing personal information with AI platforms. While Ghibli-style transformations may seem harmless, the long-term implications of handing over personal data to AI models remain a pressing concern.
As AI art generation continues to evolve, users must weigh the excitement of these viral trends against the potential risks of data misuse and privacy violations.
