OpenAI’s Efforts to Combat AI-Assisted Cheating: Text Watermarking and Beyond
OpenAI is actively researching text watermarking techniques to address concerns about academic dishonesty involving its AI language model, ChatGPT. This effort is aimed at developing a tool to detect and expose instances where students might use AI-generated essays to cheat. The update, first highlighted in a blog post from May, reflects OpenAI’s commitment to maintaining…