Why You Should Think Twice Before Uploading to Ghibli AI Generators: See The Dark Side of AI Portraits

A new obsession has taken over social media feeds: Ghibli-style portraits. OpenAI’s ChatGPT and Elon Musk’s Grok 3 have given users the ability to transform their selfies into dreamlike scenes reminiscent of Japanese animator Hayao Miyazaki’s cinematic masterpieces. At first glance, it’s harmless fun. Who wouldn’t want to see themselves reimagined with the gentle whimsy of cult movies Spirited Away or My Neighbor Totoro? But lurking beneath the nostalgia and aesthetic delight is an unspoken question: What happens to all those faces?

The Honeypot of Biometric Data

The launch of OpenAI’s Ghibli-style generator last week triggered an instant viral storm. Politicians, celebrities, and everyday users rushed to upload their photos, eager to see themselves reinterpreted through the signature brushstrokes of Studio Ghibli. Grok 3, Musk’s AI chatbot, wasn’t far behind, adding its own Ghibli-style rendering feature to xAI’s offerings. And just like that, thousands, possibly millions, of people willingly handed over fresh facial data to AI platforms.

Privacy watchdogs have taken notice. On X (formerly Twitter), digital rights activists are waving red flags, warning that these AI generators could be Trojan horses designed to amass a vast database of human faces. Unlike web-scraped data, which falls under regulatory scrutiny, these uploads are submitted voluntarily, sidestepping legal roadblocks like GDPR’s “legitimate interest” test. According to Luiza Jarovsky, co-founder of the AI, Tech & Privacy Academy, this shifts OpenAI’s data collection into a different legal category—one that allows greater freedom in processing biometric information.

“People are unknowingly handing OpenAI a goldmine of fresh, high-quality facial images,” Jarovsky wrote in a post. “Their policy explicitly states that user inputs can be used to train AI unless they opt out.”

This raises a chilling prospect: OpenAI and xAI now have access to a flood of user-provided images, many featuring clear, high-resolution faces: ideal material for refining facial recognition and deepfake technologies. Even if OpenAI and xAI claim not to store these images, what assurances do users have?

Was the Trend Engineered?

Some skeptics believe the Ghibli AI trend didn’t go viral organically. A growing theory suggests that OpenAI CEO Sam Altman deliberately enlisted celebrities and influencers to post their AI-generated portraits first, ensuring the trend would spread rapidly to regular users. By creating FOMO (fear of missing out), the company could accelerate mass adoption and, in turn, collect an unprecedented volume of facial data in record time.

This wouldn’t be the first time a tech company leveraged influencer marketing to push a new feature, but if true, it raises ethical questions. Was this a calculated move to build AI training data under the guise of a viral trend? And if celebrities were in on it, did they disclose any promotional agreements? OpenAI has not commented on these speculations, but the coordinated nature of the trend has left many questioning its origins.

Studio Ghibli’s Silent Scream

The ethical landmine extends beyond privacy. For years, Japanese animator Miyazaki, now 84, has made his disdain for AI and automation clear. In 2016, a video of him reacting to an AI-generated animation went viral... his disgust was palpable. “I strongly feel that this is an insult to life itself,” he said.

That AI models are now generating Ghibli-style images using his aesthetic without his consent would likely provoke a similar response. AI-generated art that mimics human creativity is already at the centre of an intense copyright battle, and this latest trend adds another layer of unease: Are AI tools actively eroding the legacy of master artists while harvesting user data?

ChatGPT’s Response

While OpenAI has yet to release an official statement on data safety, ChatGPT’s own response to a user’s query was revealing:

“No, it’s not safe to upload personal photos to any AI tool unless you’re certain about its privacy policies and data handling practices. OpenAI does not retain or use uploaded images beyond the immediate session, but it’s always best to avoid sharing sensitive or personal images with AI services.”

Grok 3 was more evasive about the prompt question:

“xAI doesn’t explicitly detail how long it retains uploaded images or if they’re used to train future models. Many AI tools store uploaded data on servers, where it could be vulnerable to breaches. Unless you opt out (check X settings if you’re accessing Grok via the platform), your images might also feed into improving the AI.”

So, if a user doesn’t take proactive steps to opt out, there’s a strong possibility their images are being stored and repurposed.

The Biometric Arms Race

The risk isn’t just about AI training. Biometric data is among the most valuable digital commodities. Cybersecurity experts warn that these uploaded images could be exploited in ways users haven’t considered. Once an image is processed, AI systems theoretically have the ability to generate new variations of a person’s likeness, raising concerns about identity theft and deepfake misuse.

Himachal Cyber Warriors, a cybersecurity collective, issued a stark warning: “Think before you #Ghibli. That cute Ghibli-style selfie? It might cost more than you think. Your photo could be misused or manipulated. AI may train on it without your consent. Data brokers might sell it for targeted ads. Stay cyber smart.”

How to Protect Your Privacy

As AI-generated art becomes increasingly sophisticated, users need to approach these trends with caution. Before uploading a selfie to the next viral AI tool, consider these privacy safeguards:

1. Think twice before sharing personal photos... especially high-resolution images.
2. Use anonymous or artistic avatars instead of real selfies.
3. Review AI platforms’ privacy policies and opt out of data retention whenever possible.
4. Limit permissions for apps that request camera or gallery access.
5. Avoid facial recognition for device unlocking. Use a PIN instead.

The Ghibli controversy underscores a critical point: AI advancements are outpacing public awareness of their implications. Tech companies like OpenAI and xAI frame these features as innocent and entertaining, yet they often come with invisible trade-offs.