Actually, that isn't true.
Many building code laws (and by extension hotel laws requiring fire escape access etc) were enacted because of large scale urban fires (eg http://en.wikipedia.org/wiki/History_of_fire_safety_legislat..., http://en.wikipedia.org/wiki/Great_Baltimore_Fire#Legacy).
In general, building codes and related safety regulations for hotels have strong scientific and engineering testing behind them.
It's true that these regulations are often imperfect. But it is also true that they have been tested and proven to be better than the absence of them.
However, safety, like anything else is a tradeoff. There is no such thing as a perfectly safe building. Even if you were to make it totally safe, it would be too expensive to use or construct.
You want safety? Just how much are you willing to pay?
Given that safety costs, it is usually appropriate that government regulates the minimum acceptable standard.
Beyond that, the market decides (see some active safety features in cars etc)
MGM Grand Hotel fire, 1980.
To restate: one of the justifications for regulating landlords is the possibility that they may offer unsuitable, even unsafe housing, in desperate circumstances. But where's the evidence that this is happening in the short-term rental market?
Gathering statistics for short term home/apartment rentals would be rather tricky. But apartment fires seem a frequent enough occurrence: http://cbs5.com/local/fatal.fire.firefighters.2.1823700.html... of 8 died in fire while family visiting relatives in San Jose); http://www.wtnh.com/dpp/news/fairfield_cty/dozens-displaced-...; http://www.ksl.com/?nid=148&sid=11656807; http://www.thedenverchannel.com/news/24388077/detail.html.
I have no opinion on the relative dangers of fires to residents vs short-term visitors, nor am I saying that regulation of short-term lets is necessary. I posted only to respond to the general contention that the market makes safety regulation unnecessary or superfluous.
That government licensing will protect you from these better than the market is an untested assumption.