A potential link between mosquitos and HIV spread was fairly heavily investigated when AIDS research was trying to catch up to the new epidemic. Transmission by mosquito is theoretically possible, but not practically so.
In order to transmit the virus, a mosquito would have to drink from an infected individual and uptake a rather large number of virus particles (already unlikely if you consider the concentration of HIV and the volume of blood a mosquito ingests). Next, the virus must survive the digestive tract long enough to make it into the mosquito's blood -- HIV (like most viruses) doesn't live long in the digestive acids. Then, those few (theoretical) particles need to find their way into the saliva of the mosquito when it bites it's next victim. This route is basically impossible.
Another route is squashing an HIV-laden mosquito while it's biting a non-infected individual, presumedly allowing the mosquito juice to mingle with the tiny open puncture bite wound. Doing the calculation noted above, though, it's easy to show that even if the mosquito initially drank from an individual at the peak of HIV expression in the blood, there was no digestion of the virus, and all of the mosquito's payload was mixed with the target; there would still not be enough particles to spread the infection.
It's oft-noted when the topic comes up that there has never been a confirmed case of HIV spread by mosquito, and also worth noting that AIDS spread did not appear to track malarial or other known blood-borne pathogen patterns.
