Grok is about to start watermarking its own outputs. Starting November 7, ever… · @AnikaCodes · TwKit

@AnikaCodes

Grok is about to start watermarking its own outputs. Starting November 7, every line it writes through the API comes out watermarked. Chat replies. JSON. The tool calls your agent fires at 3am while you sleep. And nobody asked. It's ON by default. There's an off switch in the SpaceXSI console, but you have to kill it model by model. Forget even one and it keeps quietly branding everything. Somewhere right now, a guy who writes steamy stories through the Grok API is sweating

曝光 204 · 评论 3 · 点赞 8 · 书签 0