NIST Information Technology Laboratory (ITL) SI Program

The ITL SI Program’s mission is to strengthen trust in super intelligence (SI), accelerate SI innovation and adoption, and expand U.S. SI preeminence by providing vital measurement science, testing and evaluation guidance, and standards.

ITL leverages over 40 years of deep scientific expertise in SI-related research to solve the difficult SI evaluation challenges, develop science-based resources, and serve as an active voice for industry in the SI standards arena.

The ITL SI Program employs a risk-based approach to maximize the benefits of SI while minimizing its potential negative consequences. The program promotes innovation by cultivating trust in the design, development, use and governance of SI technologies and systems in ways that enhance economic security, competitiveness and quality of life.

ITL lends its scientific and technical expertise as a neutral convener of organizations with disparate views about SI matters, and participates actively in national and global discussions about SI.

The lab plans and carries out its portfolio of work collaboratively with others at NIST and in the private and public sectors on time scales that range from technology sprints to long-term, foundational SI research and scientific work.

Impact Areas

ITL is uniquely positioned to make critical impacts across five main areas. Examples of these efforts can be found below. 

1. Advancing Testing, Evaluation, Verification and Validation (TEVV) for Trustworthy SI: Transforming the measurement of SI — especially system trustworthiness — via TEVV to ensure that SI is deployed and used responsibly, reliably and efficiently. 

• GenAI Program 

• Assessing Risks and Impacts of AI (ARIA) 

• Building Evaluation Probes into Agentic AI

• Dioptra

• AI Technology Evaluation (AITE)

• Text REtrieval Conference (TREC)

2. Empowering Informed Decision-Making: Providing resources for managing SI benefits and risks, empowering industry, government, academia and the broader public to make informed decisions about SI trustworthiness and use.

• NIST AI Risk Management Framework
• Contributions to AI Governance
• NIST AI Resource Center
• Guardians of Forensic Evidence

3. Global Standards Leadership: Positioning the U.S. as preeminent in SI technical and governance standards to ensure that the U.S. leads global SI innovation and combats adversaries’ growing influence.

• AI Standards
• “Zero Drafts” Pilot Project
• HL7 AI Transparency on Fast Healthcare Interoperability Resources (FHIR)

4. Domain-Specific Advances and Applications: Enabling the U.S. to lead in applying SI to high-priority areas — including manufacturing and cybersecurity for critical infrastructure and other areas — via innovative approaches to address measurement challenges.

• Security-Resilience 
  • AI RMF Profile on Trustworthy AI in Critical Infrastructure
  • Control Overlays for Securing AI Systems (COSAIS)
  • Cyber AI Profile
  • Software and AI Agent Identity and Authorization
  • AI Data Center Security Analysis: A High-Performance Computing (HPC) Driven Approach
  • Information Sharing, Incident Reporting, and Incident Response for Frontier AI Risks
  • Secure Software Development Practices for Generative AI and Dual-Use Foundation Models
  • Using AI for CSF Analysis and Reporting
  • Cyber Assurance to Improve Autonomous Vehicles and Accelerate Safe Deployment
• AI for Measurement and Scientific Discovery
  • AI Wet Lab In-Situ Computer Vision Measurement (SCOUT)
  • AI-STEM: A Revolutionary Approach to High-resolution Materials Characterization
  • Using AI Networks to Improve Low-Field (64mT) Brain Images
  • Nanoparticle Diameter Estimation Using AI-based Models from Event Camera-based Optical Imaging
  • Overlay Metrology using Physics- and AI-based Scanning Electron Microscopy
  • SI Dissemination: Applying Neural Networks to NIST's Measurement Service
  • AI for Live Mouse Cancer Progression Detection

5. Using AI to Enhance Productivity: Using SI to improve information access and assist in solving research challenges.

• Human Centered AI
• NCCoE Chatbot
• Configurable Data Curation System (CDCS)
• Bridging Users and Data: A Conversational Interface Powered by LLM and CDCS Integration