Post by @lord.luke5

OpenAI is adding an invisible watermark to ChatGPT text OpenAI says ChatGPT and Codex, its coding tool, will start putting a hidden watermark in the text they write. Right now it's only in the EU, because of the EU AI Act, a law that says AI text has to be detectable by a machine. You can't see it. It's a pattern in the word choices that a detector can pick up. OpenAI's own tests show the answers score about the same with it turned on. Not going to lie, I think this gets really interesting over the next couple years. Would you want to know if something you read was written by AI? OpenAI text watermark, ChatGPT watermark, invisible AI watermark, EU AI Act, ChatGPT EU, Codex watermark, AI text watermarking, OpenAI watermark detector, AI generated text, AI regulation, OpenAI news, ChatGPT update, AI news