OpenAI, the creator of the popular chatbot ChatGPT, has released another AI-based software tool that can classify AI-generated text. The new software release was announced in a blog post written by the company in late January, as addressed by CEO Sam Altman.
The unveiling of the tool comes after ChatGPT’s widespread popularity raised issues of misinformation campaigns and academic dishonesty.
Seeing how the chatbot can generate text in response to prompts, preparing anything from full-length essays to poetry, large school districts such as New York City have already banned the AI chatbot.
As the program is accessible for free, educators have raised concerns that students will rely on the AI’s text-generating capabilities to cheat or plagiarize.
To assuage these fears, OpenAI’s new detection tool is designed to distinguish human-written and AI-generated content. This “AI classifier” is currently in public beta mode, and OpenAI openly acknowledges that the software is far from perfect in these early stages.
Currently, the program can be tricked by human-edited AI-written content and is found to be generally unreliable for text under 1,000 characters.
“We recognize that identifying AI-written text has been an important point of discussion among educators,” said OpenAI in their blog post announcement. “[Equally] important is recognizing the limits and impacts of AI-generated text classifiers in the classroom."
The company states that it will continue to work on detecting AI-generated text.