Difficult, really difficult, I think !
In a perfect world, you would ask the AI „Tell me something about the Piranha MICV“, maybe you could additionally say „But don‘t use information from sources X, Y and Z !“ and you would get a decent report, perhaps even with at least some sources given. And all this without having to search by yourself !
Not giving the sources of an information degrades the trustability of a statement, and if it comes from an AI, the loss of trust perhaps even is bigger, than when it‘s just mentioned as „I read somewhere on the net….“ , or without a mentioned source, but without the AI mentioned either.
And to be honest, I won‘t be surprised, if even principally reliable websites, mags, or other sources, are using AI in the meantime, of course, always after cross-checking the results !
And to be honest again, I probably won‘t notice, if an information comes completely from an AI, maybe just with an asserted source … or maybe, it‘s really from that source, but that source used AI ….
I have no other idea in the moment, than to treat information from AI the same way, as from other sources : Cross check, maybe cross check again, and try to find at least some „real“ sources, as that probably could prevent at least falling victim to a hallucinating AI … though not falling victim to a source, that itself had fallen for such an incident.
No other idea, sorry !