NIST has published NIST AI 600-1, the Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile.
The profile helps organisations apply the AI RMF to generative AI. It identifies risks that are unique to, or made worse by, generative AI, including:
- Confabulation (“hallucinations”) and information integrity.
- Data privacy and leakage of sensitive information.
- Information security, such as prompt injection and data poisoning.
- Harmful bias, dangerous or violent content, and intellectual property concerns.
- Value chain and component integration risks.
For each risk it suggests actions mapped to the AI RMF Govern, Map, Measure and Manage functions. It is a valuable resource for any organisation adopting generative AI tools.
Source: NIST AI 600-1 Generative AI Profile (NIST)
Need help understanding what this change means for your organisation? Get in touch for a pragmatic, no-obligation conversation.
