A new method called 'copyright traps' has been introduced to help writers and artists determine if their work has been used in AI training data. This innovative approach involves subtly marking works so that later detection of their use in AI models is possible. The introduction of copyright traps could address ongoing concerns about the unauthorized use of creative content in generative AI. This development is seen as a potential solution to the copyright question surrounding AI training data, making it less risky and cheaper for companies to embrace generative AI. The tension between AI training and Fair Use remains a critical issue, with strong legal remedies already in place to address potential copyright violations. The story was reported by MIT Technology Review.