Last week, when I was co-hosting the Code Conference, I got to talk with Getty Images CEO Craig Peters. The generative AI boom is a direct threat to Getty in many ways. For example, the company is suing Stability AI for training the Stable Diffusion model on Getty content — sometimes clearly including AI-generated copies of the Getty watermark — without permission. Getty's answer? Its own proprietary, in-house AI tool, trained — with permission — on its own content, using a model where the original creators can get paid. Getty's put some pretty strict guardrails around it for now, but, as even Craig told us, there's still a lot of work to do. Links: theverge.com/2022/9/21/23364696… theverge.com/2023/2/6/23587393/… theverge.com/2023/9/25/23884679… theverge.com/23900198/microsoft… arstechnica.com/tech-policy/202… npr.org/2023/05/18/1176881182/s… theverge.com/2023/7/26/23808184…