>The only way someone could conclude “CBT has the same outcome as medication” from the studies you linked to would be to not read them.

Fortunately for them, that's often the case. I've seen at least a couple internet arguments with LLM-generated "sources" that didn't actually exist.