I’ve been pointing out for a long time the logical implication of how they stole their training data is that they will lose access to future data. Platforms want consent, credit, and compensation. OpenAI launched with none of those. And worse for OpenAI, that data will be made available to OpenAI’s competitors. I think Medium is the 46th biggest data set, so not huge, but still meaningful. We’ve blocked OpenAI for more than a year simply because there is zero exchange of value. Meanwhile we continue to allow Google because Google built credit into Gemini and still compensates readers with meaningful traffic.