9.dxc5? is a horrible positional blunder.
The Grunfeld for black allows white a big center, and his idea is to chip away at it, or force it to advance, and then chip away at it. That center controls a lot of nice squares, and is very desirable. 9.dxc5? by white voluntarily does what black is trying to achieve in a very bad way, and worse, it turns the Bg7 into a monster. In exchange for that, you cannot even keep the pawn as black can win it two ways. It is so bad that in 5099 games in the Mega 2020 database, no one has ever played it. In addition, the eval on Stockfish goes from about +.70 to -.67 instantly.
Black can win back the pawn with Qa5, but even stronger is 9...Nd7, and the N will have an incredible c5 square as its base. I have added notes to the following line regarding other positional advantages, so be sure to read them.
[FEN ""]
1. d4 Nf6 2. c4 g6 3. Nc3 d5 4. cxd5 Nxd5 5. e4 Nxc3 6. bxc3 Bg7 7. Bc4 O-O 8. Ne2 c5 9. dxc5? Nd7 10. O-O (10. c6 Ne5 $1 11. Qxd8 (11. cxb7 Qxd1+ 12. Kxd1 Bxb7 13. Bb3 Nd3 14. Be3 Rac8 15. Kc2 Rfd8 {There are a million ways white can play from here, but they are all losing and black is all over him.}) 11... Rxd8 {And the N gets to d3 with great effect positionally or this...} 12. c7 Rd7 13. Bb5 Rxc7 {With a huge advantage.}) 10... Nxc5 11. Qc2 Bd7 12. Rd1 Qc7 {And white has big problems down the c-file, the Bc4 is in danger (if black trades for it, the two bishops are a winning advantage in this open position), and the Ne2 is not even well-placed.}