Apple’s Image Playground app has come under scrutiny for potential bias issues, as highlighted by Jochem Gietema, the Machine Learning Science Lead at Onfido. In a blog post, Gietema shared his observations, pointing out how the AI-powered app altered skin tones and hair textures in ways that appeared to reinforce racial stereotypes.
Gietema noted in a LinkedIn post, “While experimenting, I noticed that the app altered my skin tone and hair depending on the prompt. Professions like investment banker vs. farmer produce images with very different skin tones. The same goes for skiing vs. basketball, streetwear vs. suit, and, most problematically, affluent vs. poor.”
These concerns echo broader challenges faced by AI models, which are trained on vast datasets that may carry inherent biases. Similar controversies have emerged in the past, with Google’s Gemini AI receiving criticism for comparable issues.
Apple introduced the Image Playground app as part of its Apple Intelligence suite with the iOS 18.2 update. While the company has implemented safeguards—such as restricting generated images to cartoon and illustration styles and preventing deepfake creation—these recent allegations suggest that additional measures may be necessary. The app also limits prompts containing negative words, celebrity names, or public figures to prevent misuse.
Notably, Gadgets 360 staff members did not encounter any such biases during their testing. However, if the reported inconsistencies are widespread, Apple may need to strengthen its safety mechanisms to ensure an inclusive and unbiased user experience.
