4. Malicious Actors & Misuse2 - Post-deployment

Real-world risks (Risks of using AI in illegal and criminal activities)

AI can be used in traditional illegal or criminal activities related to terrorism, violence, gambling, and drugs, such as teaching criminal techniques, concealing illicit acts, and creating tools for illegal and criminal activities.

Source: MIT AI Risk Repositorymit700

ENTITY

1 - Human

INTENT

1 - Intentional

TIMING

2 - Post-deployment

Risk ID

mit700

Domain lineage

4. Malicious Actors & Misuse

223 mapped risks

4.3 > Fraud, scams, and targeted manipulation

Mitigation strategy

1. Rapidly invest in and deploy advanced AI-integrated defenses and analytical systems, such as deep learning models and blockchain intelligence, to proactively identify, disrupt, and counter the use of AI for illegal activities, prioritizing the development of countermeasures that reduce criminal compliance with AI models. 2. Establish clear, mandatory governance policies and secure design frameworks (security by design) for all AI development and deployment, including the formation of specialized national and international taskforces to coordinate the response to AI-enabled crime and set legal/ethical boundaries for appropriate use. 3. Mandate continuous, independent post-deployment auditing and performance monitoring, coupled with systematic logging and centralized intelligence sharing among law enforcement partners regarding new criminal modi operandi and the specific AI tools being misused.