OpenAI has developed a tool to detect and mark AI writing, but it may stay under wraps: All the details

ChatGPT

OpenAI's ChatGPT is a popular tool for text generation, summarisation, and other text-related queries. It would be pretty hard to make a difference between these texts from the human-written ones. However, ever since ChatGPT began making waves online, people have been concerned about the power of AI writing when it comes to plagiarism. With AI tools, many students don't need to write their essays and reports; instead, they can plug the subject into any AI generation tool and have it written everything for them.

But now the artificial intelligence company has confirmed that it is working on a text watermarking method after the report by Wall Street Journal broke out this news. According to the firm, its text watermarking method is accurate and “even effective against localized tampering, such as paraphrasing, it is less robust against globalized tampering.”

The report suggests that this anti-cheating tool has been ready for a year, but internal discussions have delayed its release. OpenAI even surveyed with the users, and nearly a third of loyal ChatGPT users were opposed to releasing this technology. The company also says that watermarking text could hurt non-English users of ChatGPT who use the chatbot for productive tasks and translation purposes.

However, in connection with images and audio, OpenAI has recently introduced an AI detection tool using its DALL-E 3 AI model. The company has implemented its watermarking technologies on audio and visual content currently and not on text, due to the potentially more severe consequences of AI-created multimedia content.

This has also been built due to the growing cases of deepfake images and videos of popular celebrities on the internet.