@dangillmor Oh hey, libel-as-a-service! Given that they appear to be outright pretending the headline is from the site in question with no clear "we pulled this out of a robot's ass" indicator, and they know full-well that "AI" is a bullshit fountain that doesn't even have the concept of "truth", I'd say that meets the bar of "knowingly false statements", regardless of any 'fine print' disclaimers they might try.

"it only lies 30%* of the time" is an admission, not a defense... They know it's prone to spitting out falsehoods, and have done nothing to correct it. Because nothing they can do can correct it, as the bullshit-fountain is foundational to the architecture of LLMs

* Number pulled out of my ass, not the real statistic. But the point is they know that "statistically plausible output" is not always correct output.