Along With OpenAI There Is A Text Detection Tool Written By ChatGPT When ChatGPT was released in 2022, many expressed concern that students would misuse it. It was estimated that ChatGPT would be misused especially in assignments or exams. Needless to say, many incidents of students stealing cheats using ChatGPT came out. To control it, many companies started to publish AI detection tools. In particular, such tools can identify whether any content has been prepared by AI or not.
Now, OpenAI, the maker of ChatGPT itself, has revealed that it has created such a tool, which can recognize the text written by ChatGPT.
According to The Washington Post, this technology can measure content written by ChatGPT with 99.9 percent accuracy, even predicting which words (tokens) will come next. Then the AI detection tool makes some changes to the token and leaves a watermark on it. Such a watermark cannot be seen by the eye. However, the said tool identifies whether the same content is prepared by AI or not.
Although this technology has been in preparation for the past year, it is said that it was not made public after a dispute arose within the company. It has been speculated that users will decrease if the tool to find the content written by ChatGPT is removed. However, some employees within the company think that this tool should be made public to maintain transparency.
Generative AI tools like ChatGPT and Gemini answer the questions asked by the user in the style of human speech. If you send a long article and ask various questions from it, they will answer it immediately. It is said that this tool has provided great support for study and research.