That's the conclusion reached by a University of Michigan study published in the January issue of Psychology of Violence:
[Researchers] Saleem and Anderson recruited 204 participants, randomly assigning them to play one of three video games for 30 minutes. Two of the games were versions of "Counter-Strike," one with Arab terrorists and the other with Russian terrorists. The third game was a nonviolent golf game.
After playing the games, the researchers assessed participants' levels of prejudice against Arabs using direct measures such as attitude questionnaires as well as indirect measures such as drawings.
The standard tests showed that playing violent video games, even those featuring Russians as terrorists, increased anti-Arab attitudes. And the drawing test showed the same effect. Participants who played a terrorism-themed video game were more likely than those who played the golf game to draw Arabs with stereotypical traits and Arab men with weapons and angry expressions on their faces.
A sample drawing from the study: