According to Odaily Planet Daily, OpenAI has added a watermark system to the text created by ChatGPT and prepared a tool to detect watermarks. The system has been prepared for about a year, but there are disagreements within the company on whether to release it because it may harm the company's profits.

The company found that this approach was “99.9% effective” at detecting AI text when there was enough AI text, and in a survey it commissioned, “a quarter of respondents globally supported AI detection tools.”

However, OpenAI is concerned that the use of watermarks will turn off the ChatGPT users it surveyed, with nearly 30% saying they would use the software less if watermarks were implemented.

Some employees expressed concerns that tricks like using Google Translate to translate text back and forth between languages, or having ChatGPT add emojis and then remove them, could easily thwart the watermark. Still, employees agreed that the approach was effective.