Comment by conception

My current AI test. There was a BBS I was on in the 90s and there was this door game I hadn't seen anywhere else. I simply describe the BBS, where it was popular, its name, the year it was around, and the BBS game and a description of it mechanics, etc.

OpenAI and Google's Deep Research produce a very long, 100% made up report. If I question the AI on the report, they both admit they just made it up.

Claude just returns, "I couldn't find anything on the BBS or the game."