OpenAI will label AI-generated materials for EU users

The American company OpenAI announced the introduction of watermarks for the labeling of materials generated by artificial intelligence (AI) models. Chatbot developer ChatGPT took this step in order to comply with the laws of the European Union, writes sb.by.

The EU Artificial Intelligence Act obliges generative AI model providers to ensure that generated text can be identified in a machine-readable format. In the coming weeks, we will add an invisible watermark to the text output of ChatGPT and Codex in the European Union.

Watermarks do not measure a person’s contribution to a material, establish ownership or liability, identify the user, or verify the authenticity of the content. At the same time, the absence of such labeling will not 100% mean the authorship of a person.

The company stressed that it plans to focus on introducing watermarks directly in the EU, rather than spreading the practice to the global market, at least at the outset. OpenAI said it will review every part of this approach as technology and standards evolve.

The CEO of OpenAI said that the benefits of using AI technologies justify the associated risks