AI and the US Constitution

Open AI's own blog post on their AI classifier, which is the underlying technology to identify whether something's written by AI or not, only correctly identifies 26% of AI written content. So if you're trying to make any some cases, quite serious decision, are we going to accuse someone of something based on something this inaccurate? I think it's going to be a recipe for disaster, basically. And there'sgoing to be all sorts of all sorts of issues.