@davidgerard I assume you've seen the story where Meta have decided to give an AI chatbot free rein to modify accounts and then let the great unwashed chat to it, with the obvious result?

https://x.com/QuinnyPig/status/2061259665376493851

[Edit: tyop.]

RE: https://infosec.exchange/@atax1a/116665647121476321

Such lies also count as fraud by false representation. While extremely unlikely since the CPS would decide it's too much like work, this can be tried as a criminal offence which carries a prison term of up to ten years in the UK.

But just because the CPS is as useful as a fishnet condom doesn't mean that someone can lie about LLM use with impunity: they can still be pursued in civil court for the potentially quite substantial costs of auditing and excising all of the code which the sloperator falsely claimed was their own work.

An AI consultant tells Axios one of their clients recently spent half a billion dollars in a single month after failing to put usage limits on Claude licenses for employees..

The even more satisfying bit is that Anthropic would have also lost around $10 billion on the deal.

(Note that this is clearly one of those "too good to check" stories. Caveat lector.)

Bring it on!

I'd stand out in it naked to cool off, but don't want to horrify the neighbours…