BTW lads, the data quoting near negligible chances of HIV transmission - 1 in 1000, 1 in 2500 etc - is totally misleading
Those stats are based on "infected source" studies - couples where one partner has HIV and is taking antiretroviral drugs. In that scenario, the chances of transmission to the uninfected partner are extremely low because the infected partner's viral load is low.
The important thing is viral load. People who have been diagnosed with HIV and are receiving HIV treatment have a low viral load.
People who have recently contracted HIV/ do not know they have the virus, and are not receiving treatment have a high viral load. The chances of catching the virus from them is very high. You are much more likely to encounter someone in this category on a one night stand, or who is a sex worker. Such people engage in bareback sex because they don't know they have HIV. People who know they have HIV generally don't have unprotected sex with randomers as its a criminal offence to knowingly infect someone without telling them.
Look at some of the other stats and you will realise that HIV is nowhere near almost impossible to catch. There are countries in Africa with an HIV + rate of over 20% of the population. In London, 1 in 7 gay men is HIV +. These are higher risk communities, of course, but you see my point, transmission of the virus happens at a way higher rate than those <1% probability stats imply.