OpenAI is working on a tool for AI-made image recognition

According to Bloomberg, the OpenAI company is producing a new image recognition tool that will probably be released to the public soon and is very accurate in recognizing images created by artificial intelligence.

Meera Moratti, CTO of OpenAI, maker of the popular chatbot ChatGPT and image producer DALL-E, announced that the company’s new tool is 99% reliable in determining whether an image is generated using artificial intelligence. He also added that the company is testing this tool so that it can fix its bugs before the public release. Of course, Moratti did not specify when this tool will be available to the public.

Recognition of images produced by artificial intelligence by the new OpenAI tool

There are currently only a handful of tools that claim to identify AI-generated images or content. But it is said that some of these tools do not have the necessary accuracy and cannot recognize the images correctly.

For example, OpenAI introduced a similar tool in January that could recognize AI-generated text. But this tool was abandoned in July due to its unreliability. Now the OpenAI company says that it is working on improving this tool and claims that it has adopted methods that can identify images and texts produced with artificial intelligence with high accuracy.

Due to the ever-increasing growth of artificial intelligence-based tools, the need for recognition tools for content and texts produced with this technology is also increasing. It is not yet known what the next version of OpenAI will be marketed under, but it seems that the optimization of these tools is still going on inside the company.