OpenAI is to launch a tool that can identify images created by the DALL-E 3 model and transform text into images using artificial intelligence.
According to Reuters, the announcement was made on Tuesday in response to concerns about content generated by artificial intelligence.
The tool, says OpenAI, had a success rate of around 98% in identifying images generated by DALL-E 3 during internal tests. It can also handle common modifications such as compressions, cropping and changes in saturation.
ChatGPT’s creator also plans to add a watermark to digital content such as photos or audio that is difficult to remove.
OpenAI has joined a group that includes Google, Microsoft and Adobe to discuss the contours of creating digital content with artificial intelligence.
Leave a Reply