cw: WMD/ genocide
@cadadr There is of course some work in the field, though it's not given the emphasis it deserves. Several commentators have noted that CompSci, unlike physics, has not yet had its Hiroshima moment (Ex-Googler and G+ architect Yonatan Zunger, trained in physics, among them).
That notion itself may be flawed: consequences of computer-based moral failure rarely arrive as blinding insights of unignorable magnitude airdropped with precision and creating both tens of thousands of martyrs and witnesses.
The same conflict which birthed Little Boy also snuffed the souls of 6 million Jews (and others) with less haste, but punch-card precision, tabulated and enumerated by IBM as per contract operating in and for Nazi Germany, recording data with serial numbers, some of which still remain tatooed on the arms of survivors.
And yet computer science is almost wholly unaware of its Holocaust past.
There's William J. Rappaport's Philosophy of Computer Science, still in development, which includes chapters on ethics and ethics in AI specifically:
There is the startlingly prescient writing by Internet (or proto-Internet) pioneers such as Paul Baran (co-inventor of packet-based networks), writing at RAND in the 1960s on issues of ethics, morality, and social responsibility.
(Those writings are now published free of charge online at my request.)
For people who care about, support, or build Free, Libre, and Open Source Software (FLOSS).