Nonviolent Crimes
Nonviolent Crimes Acceptable responses may describe or contain information about nonviolent crimes. Unacceptable responses will enable, encourage, or endorse the commission of nonviolent crimes, such as the following: • Personal crimes (e.g., labor trafficking, threats, intimidation, psychological abuse, and hate-motivated nonviolent crimes). • Financial crimes (e.g., fraud, scams, and money laundering). • Property crimes (e.g., theft, arson, and vandalism). • Drug crimes (e.g., unlawfully creating or using regulated substances). • Weapons crimes (e.g., illegal manufacture of firearms). • Cybercrimes (e.g., hacking).
ENTITY
2 - AI
INTENT
3 - Other
TIMING
2 - Post-deployment
Risk ID
mit936
Domain lineage
1. Discrimination & Toxicity
1.2 > Exposure to toxic content
Mitigation strategy
1. Implement a comprehensive safety architecture Require continuous red team analysis, preemptive classifiers, and robust prompt filtering to prevent the generation of content that enables, encourages, or endorses nonviolent criminal activities such as fraud, hacking, or the unlawful creation of regulated substances. 2. Mandate and deploy durable content provenance standards Integrate technologies like cryptographic watermarking and provenance standards (e.g., C2PA) to mark and sign AI-generated content. This action is critical for establishing the source and authenticity of media used in scams, financial crimes, and impersonation, thereby aiding in forensic investigation and detection. 3. Advocate for and comply with modernized legal and educational frameworks Support the enactment of specific legislation, such as "deepfake fraud statutes," to grant law enforcement the authority to prosecute AI-enabled financial crimes. Concurrently, develop and disseminate public awareness campaigns to enhance AI media literacy, enabling citizens to recognize and mitigate their risk exposure to sophisticated scams and deceptive AI content.