P2863 – Recommended Practice for Organizational Governance of Artificial Intelligence (Policy)
P3142 – Recommended Practice on Distributed Training and Inference for Large-scale Deep Learning Models (Architecture)
P3193 – Recommended Practice on Large-scale Pre-trained Deep Learning Model Application Framework (Architecture)
P3419 – Standard for Large Language Model Evaluation (Evaluation)
P3396 – Recommended Practice for Defining and Evaluating Artificial Intelligence (AI) Risk, Safety, Trustworthiness, and Responsibility (Policy)
P3123 – Standard for Artificial Intelligence and Machine Learning (AI/ML) Terminology and Data Formats (Semantics)
P3394 – Standard for Large Language Model Agent Interface (Architecture)
Current Standards Needs
Establish common ground and best practices for AI governance across jurisdictions
Enable interoperability and communication for innovative and dynamic AI technologies both in general AI areas and also for vertical industries
Provide frameworks and guidelines for key areas, such as foundational models, GenAI, language models, AI agents, edge AI, explainable AI, federated learning, model evaluation, AI risk and trustworthiness, etc.
Standardize terminology, data formats, and interfaces to facilitate the integration of AI systems
Standards Stakeholders
AI technology providers and developers
Organizations deploying AI systems across various domains and vertical industries such as automobile, education, power generation, healthcare, financial services, etc.
Government agencies and policymakers
International Organizations, other standard development organizations, public interest groups, and advocacy organizations
Academia and research institutions working on advancing AI