I think that this is actually a good question that could be made better by asking are humans risk averse.
Many human Beings gamble with their health - smoking and drugs.
Some people will knowingly and willingly go to certain death - suicide bombers.
Some people will knowingly and willingly seek out dangerous thrills - gambling, mountaineering and other extreme sports.
The 20th century documents in great detail - War and violence - these are not risk averse activities.
I don't know a great deal of economic theory - but my understanding is that risk is positively correlated with gain. The greater the risk the greater the possible gain. Behavioral economics critically examines the assumption that human beings are rational agents - on the face of it this does not seem to be true. It seems to be a generally assumed axiom for the fortuitous flow of nice economic results that follow rather than one upheld by critical observation. Human beings are far more often irrational than rational.
Rationality is a key concept for developing the idea of risk-aversion rigorously. After all to be risk averse one has to coolly weigh the risks involved in decisions.
The other concept that requires critical examination is the idea of perfect availability of information to all participants - just looking cursorily at any market shows that there are huge information asymmetries. These are two solid pillars of classical economics which upon examination really are quite risible. Its fairly obvious that once one begins to look at how classical economics is constructed that it has been carefully modelled on physics.
The rational agent - compare with the rational atom which collates all its collisions and then calculates the trajectory that it must embark on.
Perfect information - the perfect fluid or perfect gas of classical 19th century physics. With no viscosity and no friction. These assumptions are made to make the physics easy. If for example the atmosphere was a perfect gas we would have no weather. By analogy, one could suppose that building this assumption in classical economics one may have nice theorems but no actual - that is real economics.
There is an old adage in programming folklore - garbage in, garbage out. There is a converse that is appropriate here - perfection in, perfection out.