I will accept that they make that claim in one place.
However, there is no backreference from LAION, and LAION explicitly states that they are not associated with similar sites:
( from https://laion.ai/faq/ )
> I found a dataset containing images while searching on the internet. What about copyright then?
> Any dataset containing images is not released by LAION, it must have been reconstructed with the provided tools by other people. We do not host and also do not provide links on our website to access such datasets. Please refer only to links we provide for official released data.
So this is probably LAIONs answer, or close enough to it that I don't feel a further need to directly contact them.
Do you have any statement or evidence that might be more specific than this blanket denial?
(Edit: specifically [ laion site:haveibeentrained.com ] returns 1 google hit, while [ haveibeentrained site:laion.ai ] returns 0 hits , Many references to haveibeentrained are found in source texts that are not particularly friendly about laion. )
(Edit2: I just gave haveibeentrained.com a random photo that I took this afternoon, and that -therefore- couldn't possibly be in the LAION dataset. It found a large number of near-matches. This could be somewhat misleading to some people).
(Edit 3: TL;DR of this thread: I have encountered several sites that have made similar claims. I suspect they can't all be telling the truth at the same time. LAION makes a blanket denial about being associated with anyone like this. Therefore my assumption now is that possibly none of them are telling the truth. To refute this assumption, we would need some sort of 2nd or 3rd party positive evidence to the contrary. This might include such things as having the specific claim be corroborated by LAION, or by a known LAION associate. I am somehow doubtful that they would actually do that here though, as that would seem contradictory to the LAION FAQ. )