OpenAI Experiments with Watermarks for ChatGPT-4o's Image Generation Model

2025-04-07 / Read about 0 minute

Author：小编

OpenAI is experimenting with novel 'watermarks' for the image generation capability embedded within the ChatGPT 4o model. This initiative may serve as a countermeasure to the growing trend of users leveraging the ImageGen model to create artwork reminiscent of Studio Ghibli's celebrated style. ChatGPT's image generation model has garnered immense attention, positioning it as one of the most sophisticated multimodal AIs currently available. By implementing watermarks, OpenAI seeks to assist users in distinguishing AI-generated content. These watermarks encompass both invisible metadata and visible CR symbols, empowering users to trace the image's origin and the specific AI tool utilized via a dedicated website.

Previous page：Tang Yu, Li Dongsheng, and Colleagues from Nationa...

Next page：42 A-share Companies Unveil First-Quarter Earnings...

Return to List

Hot Reading

2 day ago

Amazon Fire TV’s new interface is now rolling out in the US

2 day ago

Cohere launches a family of open multilingual models

2 day ago

Apple is reportedly cooking up a trio of AI wearables

2 day ago

Apple Podcasts is getting an enhanced video experience this spring