There is no attack to "refute", and 2.c4 is just a blitz move.
Just take on d4, when white probably has to take back with the queen (as ...Bb4+ can become annoying) and then normal development with ...Nc6, ...Nf6, ...Bb4 and so on where Black is very comfortable (in an approximately equal position).
An even better idea is to stop playing 1...e5, unless you have certain S/M tendencies.
I usually play the Charlick Gambit against d4 (1.d4 e5 2.dxe5 d6 3.exd6 Bxd6), but recently players have been answering c4 instead. I've played three games against d4 in the past two days, and 2.c4 was played in all three. I tried to research it, but to no avail; Chess.com and lichess both think it's out of the book. So what's going on? Are people just premoving the Queen's Gambit, or is there some viral opening trap going around?
Second question: what's the right response from Black? Nf3 transposes to the Budapest Gambit, but the explorer says that line favors White. The explorer claims 2... exd4 3.Qxd4 Nf3 favors Black, but only records 15 games in that position. I've been trying Bb4+ in the style of the Bozo Indian, with some success, but I'm not confident it truly refutes White's attack.