I was working at CERN bubble chamber experiments back in the early seventies. The accelerators required a vacuum so as to be able to sustain the beam which makes many turns in the circle ( practically velocity of light), so the best possible vacuum is and was a necessity.
After the generation of the beams , the beam lines did not need a vacuum because the probability of scattering in air is very small, and mainly a bit of ionization can happen. Even in the dense liquid of the bubble chamber maybe one in ten incoming particles interact; the intensity was low, controlled to be ten or so particles at a time, for clear pictures :
This is an antiproton beam entering on the left.
When we got a glitch in data taking we would complain that the cat had entered the beam line ! Yes, a person could get in the beam. There were physicists in the early years who would center the beam by the Cerenkov light it made passing trough the eye. I know of one who died of cancer of the retina.
In those early times a proton beam could be a primary beam steered towards an electronic detector. In this case the beam would have many more particles and would present a greater danger. Still it could go through the air, depending on the experimental setup, may be for special checks with low intensity, but one should be careful of the radiation induced and certainly not to enter the beam line.