I’m guessing that a clinical trial like this needs to be run in an at-risk population, and that population can’t be taking another preventative medicine. Finding people who will take a vaccine but won’t take a pill sounds possible but difficult.
As I understand it, existing HIV medications need to be taken consistently, forever, in order to be effective, and there's still a (really tiny but extant) chance of still being contagious? Uncertain on that last part.
However, you still need the drugs constantly once you've been infected. If a vaccine could be provided that eliminated the need for sustained medication (and thus the risk of loss of access, which can be pretty high in some African countries, I'd imagine), you would have a treatment that objectively is better for someone's quality of life.
So I guess you'd ethically be weighing the risk of misbehavior of the experimental medication causing damage to life versus the long term, sustained risk of loss of access and associated damage caused by just sticking with and improving the existing treatments.
I'd much rather see a population given a one shot treatment than to be permanently beholden to the whims of a pharmaceutical company. Sure, they're BOTH from the same company, but one's scarier than the other in my mind.
The ethical issue is that to study people taking an unproven vaccine, you do have to specifically have participants not take an existing well proven prevention medicine. Since existing prep works well enough that you would never be able to determine vaccine efficacy if they were using it. Having a vaccine is still strictly better, but that's the issue, not comparing to lifetime treatment.
To your uncertainty, effective treatment (reducing viral load to undetectable) reduces transmission to zero (https://www.niaid.nih.gov/diseases-conditions/treatment-prev...).
That does kind of muddy the waters a bit. I think it's better for everyone to have more treatment options, in this case, and a vaccine is certainly better, but ... deciding who has to take the hit in the risk department to develop one is odd.
It still feels like it should be ethically possible. Logically, we don't want to pigeonhole ourselves into something of a "local minimum" when it comes to a viable solution to a problem, IE a situation where we have something that kind looks better than the surrounding avenues of research and the energy investment needed to find a lower-risk, higher-reward outcome is too high to be worth bothering.
People are also much more willing to take one or a few doses for free vs having to continuously take a preventative forever (even if free there are time and transport costs to getting refills).