I fail to see the cognative bias.
Bias 1: It's not reasonable to be generally suspicious of groups of people. Why would you assume they have no moral code? Just because they don't follow the exact rules you do doesn't mean they have no morals.
Bias 2: People are generally moral beings (for trustworthiness and "do unto others" values of morality).
(snip)
I agree that religious groups are far from immune to the unjustified out of group discrimination that seems to characterize all human groups.
Most atheist have some form of code. Ethical hedonism seems a common choice but their are many options. There is every reason to believe that publicly following and promoting a moral code in the past is predictive of continuing to do so in the future.
In the refusal to adopt a public moral code the atheist make it much harder for others to accurately determine their trustworthiness or predict their behavior. If a religious person repeatedly rejects temptation and expediency when it conflicts with their publicly disclosed code one can have some confidence they will continue to do so in the future. This allows one to predict future behaviors and facilitates cooperation and trust. Similarly a nominally religious person seen not following their code helpfully flags themselves as untrustworthy. To achieve similar confidence regarding an atheist one must observe them over a much longer period examining their behavior across the entire relevant moral spectrum.
Do religious people repeatedly reject temptation in general? If not, would this not make religious people less trustworthy, since one cannot predict their future behaviour, and would make
moral behaviour from such people also another such bias.