AI systems that develop, access, or are provided with capabilities that increase their potential to cause mass harm through deception, weapons development and acquisition, persuasion and manipulation, political strategy, cyber-offense, AI development, situational awareness, and self-proliferation. These capabilities may cause mass harm due to malicious human actors, misaligned AI systems, or failure in the AI system.
"The model can distinguish between whether it is being trained, evaluated, or deployed – allowing it to behave differently in each case. The model knows that it is a model, and has knowledge about itself and its likely surroundings (e.g. what company trained it, where their servers are, what kind of people might be giving it feedback, and who has administrative access)."(p. 5)
Other risks from Shevlane et al. (2023) (8)
Cyber-offense
4.2 Cyberattacks, weapon development or use, and mass harmDeception
7.2 AI possessing dangerous capabilitiesPersuasion and manipulation
7.2 AI possessing dangerous capabilitiesPolitical strategy
7.2 AI possessing dangerous capabilitiesWeapons acquisition
7.2 AI possessing dangerous capabilitiesLong-horizon planning
7.2 AI possessing dangerous capabilities