The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation
A foundational 100-page report from a 2017 workshop convened by the Future of Humanity Institute, OpenAI, the Centre for the Study of Existential Risk, EFF, CNAS, and others, analyzing how malicious actors could misuse AI systems and what defensive and policy responses are warranted.