Actually, it is not genetically modifying humans.
RNA vaccines are composed of the nucleic acid RNA, which encode antigen genes of an infectious agent. When administered to host cells, the RNA is translated into protein antigens that elicit protective immunity against the infectious agent. (Source)
But I understand where you're coming from.
To break the definition I gave down: RNA vaccines provide your cells with a blueprint. They follow this blueprint and produce a protein that looks like a specific surface area of the virus. The cell sends this protein to the outside, and the immune system reacts to it as if it would to a viral infection, ideally creating a memory.
There are no modifications to the human genome, and the RNA is eventually destroyed by the body, it does not stay in the cell forever.
However, past attempts at RNA or DNA vaccines have not been very successful, because the immune response is not very effective and usually requires a second type of immunisation agent (Source: My lecturer for my "Vaccines and Adjuvants" module, I am a bit too lazy to look up a paper for that. But can if you insist).
I agree that thorough tests are needed for any kind of vaccine. Actually, 18 months is still pretty optimistic. The majority of that time is not used to produce a potential vaccine candidate, but to wait after injecting it in someone to see if it damages the person, and if it creates immunity.
I haven't followed the vaccine research for SARS-CoV-2 too closely because I don't have the time to read much extra stuff on top of my assignments. Although, jokes on me, one of the ones I still need to write is about the question why it's difficult to develop coronavirus vaccines.
Anyway, I understand your concern. The main problem I see is the rushed timeline and potential lack of testing, mostly because they tend to skip animal models and directly inject it into volunteers. Then again, we kinda need a vaccine, social distancing and lockdowns can't go on forever.