OpenAI, the creator of the popular chatbot ChatGTP, has released a software tool to identify text generated by artificial intelligence, the company said in a blog post on Wednesday.
ChatGPT is a free program that generates textin response to a prompt, such as articles, essays, jokes and even poetry. It has gained popularity since its debut in November, while raising concerns about copyright and plagiarism.
The AI classifier, a language model trained on a dataset of pairs of human-written and AI-written text on the same topic, aims to distinguish text written by AI. The company said it uses a variety of providers to address issues such as automated misinformation campaigns and academic dishonesty.
In its public beta mode, OpenAI acknowledges that the detection tool is very unreliable on texts under 1,000 characters, and AI-written text can be edited to trick the classifier.
OpenAI said that the classifier will be publicly available to get feedback on whether imperfect tools like this one are useful.
We recognize the limits and impacts of AI-generated text classifiers in the classroom, and we recognize that this is an important point of discussion among educators. Some of the largest U.S. school districts, including New York City, have banned the AI chatbot because of concerns that students will use the text generator to cheat or plagiarize, after ChatGPT debuted in November.
Other third-party detection tools, including GPTZeroX, have been created by others to help educators detect AI-generated text.
OpenAI is working on the detection of AI-generated text by engaging with educators to discuss ChatGPT's capabilities and limitations.