That would imply that producing ethanol requires more fuel than it produces, so producing ethanol costs more money (for fuel) than the resulting fuel can be sold for - which is simply not true.

While you might argue that some cases of bioethanol in the first world are essentially just farm subsidies, producing sugarcane ethanol in favorable conditions (e.g. Brazil) definitely is profitable in the financial sense and also in the CO2 sense because it saves most of CO2 emissions compared to the fossil fuels they replace.

Of course, even in ideal conditions any fuel has zero net emissions and can't possibly have a net negative CO2 contribution by itself (that would require carbon sequestration) but every liter of gasoline/diesel that gets replaced by a low-CO2 fuel means a reduction of CO2 emissions.

You shouldn't compare something against a fictional zero, you compare it against realistic alternatives, so CO2 emissions are not only acceptable but actually desirable if they replace much larger CO2 emissions.