The White House has announced the completion of several key actions under the AI Executive Order. The actions aim to ensure the safe, secure, and trustworthy development of artificial intelligence. As part of the 270-day milestone since the EO was signed, federal agencies have released guidance documents and reports.
These support AI governance and risk management. NIST has published the AI Risk Management Framework (AI RMF 1.0). It focuses on generative AI (GenAI).
The GenAI Profile assists organizations in managing risks unique to GenAI. It outlines twelve specific risk areas, including:
– Chemical, biological, radiological, or nuclear weapons capabilities
– Confabulation (e.g., hallucinations or fabrications)
– Dangerous or hateful content
– Data privacy concerns
– Environmental resource utilization in training models
– Harmful biases
– Human-AI interaction configurations
– Information integrity and security
– Intellectual property considerations
– Obscene, degrading content
– Value chain and component integration
The document provides over 200 actions for companies. These help govern, map, measure, and manage these risks effectively.
NIST also published a companion to its Secure Software Development Framework (SP 800-218).
Key ai actions completion update
It focuses on assisting developers in securing software involved in AI models and systems.
This is particularly for those used for dual purposes. Additional key actions were taken by various federal agencies and departments. These include the Chief Data Officers Council, Department of Commerce, Department of Defense and Homeland Security, Department of Education, Department of Energy, Department of Justice, Department of State, Department of Veterans Affairs, National Science Foundation, and National Security Council.
NIST also issued final guidance on global AI standards. It emphasizes international cooperation for alignment in AI governance. This includes strategies to prevent misuse of AI technologies and promote best practices globally.
The White House’s approach underscores the importance of cross-sectoral collaboration and international engagement. The newly released guidance documents and frameworks set a foundational structure. This helps organizations navigate the complexities of AI risk management.
It paves the way for more secure and ethical AI innovations.
