Here comes one of my occasional blog-like replies that may be useful in response to future questions. Maybe more detail than you want or need, but here goes. Potential risks for HIV were first outlined 30-40 years ago, when the worldwide HIV/AIDS epidemic was new but knowledge was still evolving, i.e. before clinical experience matured and careful research could be designed and conducted. Many of those initially believed routes of infection proved to be rare or non-existent. Human nature being what it is, and some agencies (especially non-scientific ones) having moralistic or political motives, some zero or near-zero risk exposures commonly show up with no qualifications to define actual risks, case numbers, etc. Human fears and anxieties also play roles and it's always best to avoid websites by and for people with a particular health problem or fears of it -- like Reddit, for example -- and stick with professionally run or moderated sources. And there's simply the fact of efficiency:  It takes a lot fewer words (e.g. on brief statements or written hand-outs) to just say (for example) that oral sex can transmit HIV, without describe risk in quantitative or statistical detail, or the comparative risks versus other kinds of exposure. A particularly egregious error made by many otherwise responsible agencies is to list "blood exposure" without qualitifcation, leading people to believe that merely touching a spot of dry blood might be risky, when in fact there are almost no blood-transmitted HIV infections other than those from shared injection equipment, despite many people's occasional interaction with other persons' blood.