OpenAI Developing New AI Detection Tool with 99% Accuracy in Identifying AI-Generated Images
-
News: OpenAI recently announced it is developing a tool designed to detect AI-generated images with high accuracy. The tool boasts a reliability rate of up to 99%, as confirmed by OpenAI's Chief Technology Officer Mira Murati. This announcement was made at the Wall Street Journal's Tech Live conference.
Image credit: AI-generated image, licensed by Midjourney
Mira Murati stated that OpenAI's tool is already undergoing internal testing and is planned for public release in the future. Although there is no specific timeline for its release, the technology's potential is highly promising.
While there are already tools claiming to detect AI-generated images or other content, they often suffer from inaccuracies. For example, OpenAI released a similar tool earlier this year to determine whether text was generated by AI, but it was shelved in July due to instability. However, OpenAI has stated that they are improving the software and are committed to developing methods to detect whether audio or images are AI-generated.
The need for such detection tools is becoming increasingly important, as AI tools can be used to manipulate or create fake news reports about global events. Additionally, Adobe Inc.'s Firefly image generator is also addressing this issue by promising not to produce content that infringes on creators' intellectual property rights.
OpenAI executives also revealed that they will introduce a new AI model following GPT-4. Although OpenAI has not publicly announced the name of the model succeeding GPT-4, they filed a trademark application for "GPT-5" with the U.S. Patent and Trademark Office in July this year. This news has garnered significant attention in the industry, showcasing OpenAI's continued innovation in the field of artificial intelligence.