The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation
Artificial Intelligence
2024-12-03 v2 Cryptography and Security
Computers and Society
Abstract
This report surveys the landscape of potential security threats from malicious uses of AI, and proposes ways to better forecast, prevent, and mitigate these threats. After analyzing the ways in which AI may influence the threat landscape in the digital, physical, and political domains, we make four high-level recommendations for AI researchers and other stakeholders. We also suggest several promising areas for further research that could expand the portfolio of defenses, or make attacks less effective or harder to execute. Finally, we discuss, but do not conclusively resolve, the long-term equilibrium of attackers and defenders.
Cite
@article{arxiv.1802.07228,
title = {The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation},
author = {Miles Brundage and Shahar Avin and Jack Clark and Helen Toner and Peter Eckersley and Ben Garfinkel and Allan Dafoe and Paul Scharre and Thomas Zeitzoff and Bobby Filar and Hyrum Anderson and Heather Roff and Gregory C. Allen and Jacob Steinhardt and Carrick Flynn and Seán Ó hÉigeartaigh and SJ Beard and Haydn Belfield and Sebastian Farquhar and Clare Lyle and Rebecca Crootof and Owain Evans and Michael Page and Joanna Bryson and Roman Yampolskiy and Dario Amodei},
journal= {arXiv preprint arXiv:1802.07228},
year = {2024}
}