OpenAI Adds Invisible Watermark to ChatGPT Text Output in the EU
OpenAI announced it plans to add an invisible watermark to some text generated by its AI models, including ChatGPT and Codex, within the European Union. The company stated that this system will be applied to eligible outputs over the coming weeks as part of a broader effort to address concerns around content authenticity and misuse. The move aligns with increasing regulatory pressure in Europe regarding artificial intelligence transparency and accountability. By embedding invisible watermarks, OpenAI aims to help identify AI-generated text without altering its appearance or usability for end users. This development could influence how regulators assess compliance with AI-related laws such as the EU’s proposed Artificial Intelligence Act. The implementation is expected to be gradual but signals a proactive stance by OpenAI in adapting to regional regulations and public expectations around responsible AI deployment.
Topics: ChatGPT EU regulations OpenAI watermarking