Google Resumes AI Image Generation Feature

Google is reintroducing its AI-powered image generation feature, Gemini, after addressing significant issues that led to its suspension earlier this year. The updated Imagen 3 model will again allow the generation of images of people but with strict safeguards against producing photorealistic images, identifiable individuals, minors, and violent content. This decision follows backlash over historically inaccurate and racially insensitive images created by the chatbot, prompting an apology from CEO Sundar Pichai. Google has committed to continuous improvement and user feedback, implementing better training data to mitigate biases, including a previously noted failure to accurately depict white individuals. The new feature will be gradually rolled out to English-language users subscribed to the Gemini Advanced product, with business access coming soon. Despite improvements, Google acknowledges that the system is not perfect and will monitor its use closely.