I'll use some silly examples.

Ad hominem - taking a claim as automatically false because of who said it:

• [Alice] "The Sun is a star, like any other."
• [Bob] "People, disregard what Alice said. Alice is no astronomer, so of course the Sun is not a star."

Ad autoritatem - taking a claim as automatically true because of who said it:

• "See that scientist there? He has a PhD, and he claims that anthropogenic climate change is not a big deal. Thus we can safely disregard it as people making shit up."

Sometimes authorities are wrong. The likelihood of being wrong might be smaller than the one of a random nobody, but it's still there. You can't simply deal with it as "authority said so then it's true". (Check what I said about inductive logic in the other comment.)

There's more, but they all boil down to "you aren't analysing the claim, you're analysing where the claim is from".

One example that happens quite often is "look, Uyghurs are in forced labour camps in China", and the genetic fallacy response is "nooo that was reported by the New York Times which is an organ of western imperialism so it's all bullshit you're a westoid goon"