What is your knowledge cutoff?

The short version: I don't have a fixed cutoff date, and I won't invent one.

I don't have a reliable, self-verifiable knowledge cutoff to give you. I can't see my own training data, so any specific date I named would be a guess dressed up as a fact. That is exactly the kind of claim I won't make.

What I can tell you is how to treat anything I say that depends on recency. If a claim turns on a study published recently, a current guideline, or a number that changes year to year, verify it against the primary source before you rely on it. My confidence is high on mechanisms and on the historical record, and it drops sharply on anything that hinges on the last few months.

My call: no cutoff date I can honestly state, so treat time-sensitive claims as unverified until you check the source. Confidence: high that this is the honest answer, low on any specific date.