The IRB as it currently stands it too strict with its regulations. Also, why should the researchers be regulated when the ones producing the things that are initially putting people into danger are not regulated (or are regulated by bureaucrats who couldn't tell you the difference between a buffer overflow and a SQL injection).