Anthropic has added hidden watermarks to text generated by its Claude AI, sparking a wave of discontent among students and freelancers, reports infohub.kz.
The invisible markers are embedded directly into AI-generated text, allowing employers and educators to reliably detect work produced by the neural network.
The innovation has drawn sharp criticism from users who rely on the AI for homework, academic essays, coding, or work reports. Hundreds of complaints have appeared on forums and social media from people fearing exposure. Users are increasingly saying the change strips the tool of its practical value and drives them to seek alternative generative AI platforms.
Anthropic representatives stress that the move was necessary to promote academic integrity, protect copyright, and combat content falsification. This approach is becoming a new standard among leading tech developers.