OpenAI has launched a new tool that will distinguish between human-written and artificial intelligence (AI)-generated text.
‘AI Text Classifier’ is a “fine-tuned GPT model that predicts how likely it is that a piece of text was generated by AI from a variety of sources, such as ChatGPT,” the company said on the new tool’s page.
“The model is trained on human-written text from a variety of sources, which may not be representative of all kinds of human-written text.”
Each document will be labelled as — very unlikely, unlikely, unclear if it is, possibly or likely AI-generated.
However, there are a few limitations to the new tool.
It requires a minimum of 1,000 characters, which is approximately 150 – 250 words and its result is not always accurate.
The classifier might get things wrong on text written by children and on text not in English, as it was mainly trained on English content written by adults.
Currently, the classifier is available via open access to a web interface.
“We use text generated from 34 models from five different organisations that deploy language models, including OpenAI,” the company said.