Social media is currently overflowing with Ghibli-style portraits after OpenAI introduced ChatGPT’s Ghibli-style AI image generator. In the trademark style of Ghibli icon Hayao Miyazaki, millions of users—including politicians and celebrities—are sharing their artificial intelligence-generated images. Elon Musk’s AI chatbot Grok also has the ability to create images in the style of Ghibli. With Grok 3’s capability, users may create Ghibli-inspired graphics for free. But do you know about the dangers of Ghibli art AI?
The most recent online fad has now generated a contentious debate. Advocates for digital privacy warn that OpenAI might be secretly collecting a lot of personal data. It’s possible that users are unintentionally providing OpenAI with a sizable, excellent, and varied dataset of human faces.
What dangers come with using your own photos to create AI-generated images?
OpenAI’s Ghibli-style AI art generator alarmed digital privacy groups on X. They claimed that OpenAI is collecting thousands of individual photos for AI training by leveraging this trend.
Critics caution that users are unwittingly giving OpenAI new facial data, which could jeopardise their privacy, even as they are enjoying the feature and the newest trend.
1. Bypassing legal restrictions
Activists claim that OpenAI’s data collection method is more than just a copyright problem for AI. Bypassing the legal limitations that apply to web-scraped data, it enables the business to obtain photographs that have been voluntarily contributed.
OpenAI must provide a “legitimate interest” justification for image scraping from the internet in order to comply with GDPR (General Data Protection Regulation) requirements, which call for extra security measures to preserve user privacy.
The agreement that users provide when they upload their own photographs, however, completely circumvents these limitations and grants OpenAI greater processing freedom. Thus, OpenAI retains ownership of the originals while receiving free, high-resolution photos for AI training.
Luiza Jarovsky, co-founder of the AI, Tech & Privacy Academy, said in a thorough X post that when individuals freely upload photographs, they provide OpenAI permission to process them (Article 6.1.a of the GDPR). The legitimate interest balancing test is no longer relevant because OpenAI has more freedom under this new legal basis.
🚨 Most people haven't realized that the Ghibli Effect is not only an AI copyright controversy but also OpenAI's PR trick to get access to thousands of new personal images; here's how:
To get their own Ghibli (or Sesame Street) version, thousands of people are now voluntarily… pic.twitter.com/zBktscNOSh
2. OpenAI now has easy and free access to private photos; only they can view the originals.
OpenAI receives both the raw and modified data, but users can only view the finished “Ghibli-style” version. It may be strengthening its capacity to train AI models in the future.
In addition to the legal benefit, OpenAI is getting access to fresh, original photos, such as family and intimate photos that are shared online. OpenAI retains the original uploads, in contrast to social media corporations that have access to the AI-generated “Ghiblified” versions.
3. Your photos might become the Next AI Training Tool
Users have been warned about the dangers of posting their faces to AI platforms by cybersecurity organisations such as Himachal Cyber Warriors. According to a widely shared tweet on X (previously Twitter),
“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. Your privacy matters.”
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. Your privacy matters.… pic.twitter.com/aEjT3sHtTN
— Himachal Cyber Warriors (@hpcyberwarriors) March 29, 2025
According to OpenAI’s own privacy statement, unless users choose not to, the company gathers personal information they provide to train its AI models. Few people, meanwhile, take this into consideration when engaging in trends like Ghibli-fication.
4. Why is it risky for AI to gain your Visual Data?
According to Yann LeCun, Chief AI Scientist at Meta, for AI models to achieve human-level intelligence, visual input is necessary. According to him, texting is insufficient on its own.
He revealed that 20 trillion words—the total amount of publicly accessible content on the internet—are used to train the biggest AI models available today. However, in just four years, a child processes the same quantity of visual information.
AI must therefore process enormous volumes of visual input in order to achieve human intelligence. Getting it straight from users is the simplest method.
LeCun explains,
“We’re never going to get to human-level AI by just training on text. We’re going to have to get systems to understand the real world, and understanding the real world is really hard.”
Proton, a website that supports online privacy, mentioned on X,
“Once you share personal photos with AI, you lose control over how they are used since those photos are then used to train AI.”
The group also warned on X that,
“data could be used for personalised ads and/or sold to third parties”.
Think this is a fun trend? Think again.
While some don't have an issue sharing selfies on social media, the trend of creating a "Ghibli-style" image has seen many people feeding OpenAI photos of themselves and their families.
5. Experts also warn about the possible risk of one’s images being misused by AI for commercial exploitation
They think that people’s posted images could be utilised to create false or disparaging content.
British futurist, journalist, and AI ethics advocate Elle Farrell-Kingsley cautioned against utilising personal photos for these Ghibli-style portraits. On X, she wrote,
“Uploading photos/thoughts to AI tools risks exposing metadata, location, even sensitive data — especially for kids.
“If it’s free, you (& your data) are the price. If you’re fine with that, great but it’s good to be aware”
Experts also noted that, unlike a password, a compromised facial identity cannot be changed, so it is permanently revealed.
6. Copyright issues
According to legal experts, OpenAI’s creation of pictures that resemble Studio Ghibli films does not technically violate any regulations. The style is not specifically protected by copyright rules, according to Evan Brown, an intellectual property attorney at the US-based legal company Neal & McDevitt, who spoke to TechCrunch.
Although “style” cannot be protected by copyright, Josh Weigensberg, a lawyer at the law firm Pryor Cashman, was cited as saying that certain aspects of an artwork may violate copyright regulations if they are too close to the original.
Hayao Miyazaki, the creator of this style, dislikes artificial intelligence and picture generation. Hayao Miyazaki, a co-founder of Studio Ghibli, explained AI-generated art in a 2016 video as,
“insult to life itself.”
He added,
“I am utterly disgusted,”
“If you really want to make creepy stuff, you can go ahead and do it, but I would never wish to incorporate this technology into my work at all.”
I remember this clip of the Ghibli founder Hayao Miyazaki documentary when he says AI generated art is an “awful insult to life”.
ChatGPT responded as follows when we enquired about the safety of uploading private images to the Ghibli art generator:
Uploading personal and family photos to AI tools like ChatGPT’s Ghibli-style art generator involves certain privacy and security risks. Digital privacy experts have raised concerns that such platforms may collect and utilize uploaded images to train AI models, potentially compromising user privacy.
Thus, OpenAI’s driving force behind popular visual trends like the Ghibli generator might provide a free means of obtaining millions of photos.
Think about the risks before following the latest trends in AI images. Here are some actions you may do to safeguard your privacy and biometric data:
Do think twice before submitting your own photos for AI-generated images.
Since high-resolution photos can be scraped for AI training, it is best to avoid sharing them on social media.
Use PINs or passwords to unlock devices rather than facial recognition.
Check whether apps are authorised to use the camera to limit access.
Did you also follow the Ghibli craze? If so, after reading about the dangers, you ought to give it another thought. Feel free to express your opinions in the comment area.
