High-impact misuses and abuses beyond original purpose

Sub-category

"Since general-purpose AI systems have a large repertoire of capabilities, mali- cious actors such as foreign actors can use such systems to cause large damage if they gain unrestricted or unmonitored access to those AI systems."(p. 45)

Entity— Who or what caused the harm

Human

Due to a decision or action made by humans

AI system

Due to a decision or action made by an AI system

Other

Due to some other reason or is ambiguous

Intent— Whether the harm was intentional or accidental

Intentional

Due to an expected outcome from pursuing a goal

Unintentional

Due to an unexpected outcome from pursuing a goal

Other

Without clearly specifying the intentionality

Timing— Whether the risk is pre- or post-deployment

Pre-deployment

Occurring before the AI is deployed

Post-deployment

Occurring after the AI model has been trained and deployed

Other

Without a clearly specified time of occurrence

Supporting Evidence (1)

"For example, LLMs have strong source code generation capabilities, which can also be used to aid the creation of malware or the discovery of exploits in existing software and hardware. Relevant AI systems are ones that can help to create or can be used as weapons, such as lethal autonomous weapons (through object and movement detection and maneuvering), bioweapons (by enabling non-specialists to perform specialist lab work) [204], cyberweapons (through finding and creating exploits), and other military technologies."(p. 45)