Question being, are these a secret trove of AI tools that don't have any potential risk of future copyright claims for having been trained off of content on the internet? I feel like we're still waiting for an AI-based company to hit a landmine on that topic.