OpenAI is evolving fast by training its models on public data. Recently, it has become trendy to generate Ghibli-style photos. Hayao Miyazaki, the legendary Studio Ghibli founder, famously dismissed AI-generated art as “an insult to life itself.” Yet, OpenAI’s latest ChatGPT image generator can now create Ghibli-inspired images on demand. The problem with OpenAI is that it uses people’s data to train its models. In most cases, it doesn’t ask for permissions.
AI models like ChatGPT and its image generator rely on massive datasets to function. That includes scraping the web for articles, images, and videos. Even now, every prompt, photo, or artwork you upload can be used to refine OpenAI’s technology. As the company explains, “Training on images reflecting a vast variety of image styles allows the model to create or transform images convincingly.”
But this also raises a big question. Do people want their personal data to be used by OpenAI for training its models? We are uploading our photos to ChatGPT to generate Ghibli-style photos. It automatically gives ChatGPT permission to use the photo for training its GPT models. If you do not want to contribute your data to OpenAI’s model training, here’s how to opt out.
Stop OpenAI From Using Your Images to Train AI
OpenAI does offer a way to limit its access to your data, but you need to manually disable a few settings.
Open ChatGPT and click on your profile picture
Go to Settings and then Data Controls. Tap Improve the model for everyone. Here, disable the toggle in front of “Improve the model for everyone.”
If you use voice or video search, make sure the toggles for voice and video recordings are also disabled.
The option is also available in the ChatGPT mobile app, and the steps are the same. So, don’t forget to disable the toggle.
AI tools are exciting, but privacy concerns shouldn’t be ignored. If you value control over your data, take a few minutes to adjust your settings and prevent OpenAI from using your images and prompts to train its AI. While it will not stop AI companies from scraping public content, it at least ensures your direct contributions remain private.
Related Posts
