Ever since the launch of ChatGPT and its rivals, opinions on the AI chatbot have been divided. While some see it as important and time-saving, some others are concerned about people using AI to spread misinformation and plagiarize work. OpenAI, the company behind viral AI bot ChatGPT and text-to-image generator DALL-E, has released a new AI Classifier tool for identifying AI-written text. In an official press release, OpenAI describes how the classifier was trained, its limitations and other metrics. Check out all the details below.

In a recent press release, the company described how this new AI Classifier tool is meant to help people identify AI text. The tool does this by taking input from users and then running it through its trained data to classify it as AI or human-generated content.

AI Classifier is a language model and has been trained on a dataset consisting of pairs of human and AI written text on the same topic. OpenAI notes that human data has been collected from various sources that it believes are written by humans. This text has been split into prompts and responses to which the AI text was generated to collect a complete data set.

Although the new tool is supposed to make our lives (especially teachers and professors) easier when it comes to AI, the company has clearly stated that the AI Classifier is not completely reliable and prone to errors. In the company's internal evaluations, the classifier correctly identified 26% of AI-written text as "likely AI-written."