In response to Executive Order 13859, on August 9th, 2019, the National Institute of Standards and Technology (NIST) released U.S. Leadership in AI: A Plan for Federal Engagement in Developing Technical Standards and Related Tools.
In February of 2019, Executive Order 13859 instructed NIST to develop a plan that introduced international standards that protected federal priorities for innovation, public trust, and public confidence in systems using AI and ensured that technical standards reduced vulnerability to attacks.
After reviewing comments from over 40 organizations in the industry, academia, and the government, NIST has released its plan recommending the federal government advance U.S. AI leadership by bolstering AI standards; promoting focused research regarding trustworthiness; expanding public-private partnerships; and engaging internationally.
It specifically recommends that the federal government “commit to deeper, consistent, long-term engagement in AI standards development activities to help the United States to speed the pace of reliable, robust, and trustworthy AI technology development” (4).
To read the comprehensive plan, click here.
The plan identifies nine areas of focus for AI standards:
Concepts/Terminology
Data/Knowledge
Human Interactions
Metrics
Networking
Performance Testing/Reporting Methodology
Safety
Risk Management
Trustworthiness
Trustworthiness is an area of focus on which NIST places emphasis, as it is only now being considered by standards developing organizations (SDOs). According to NIST, our ability to understand and evaluate the decisions made by AI systems and measure their trustworthiness is limited (8). AI systems make decisions based on data-driven models; but as technology advances, further scientific testing must be done to ensure the trustworthiness, accuracy, reliability, resiliency, objectivity, security, safety, and accountability of AI systems.
The plan emphasizes that U.S. engagement in developing AI standards is imperative in strengthening the nation’s innovativeness and competitiveness, warning that a lack of involvement may pose a disadvantage for U.S. government agencies and U.S.-based companies in the marketplace.
In addition, NIST outlines steps for federal agency engagement in AI standards, including:
Identification of how AI technologies can be used to further the agency’s mission
Knowledge of existing statutes and policies regarding the participation in the development of standards
Identification of standards and related tools in need of development
Use of appropriate standards
Engagement in the development of standards to meet agency requirements
Both the Department of Transportation and the Food and Drug Administration have already begun to determine the AI standards needed for their agencies, and NIST suggests that others do the same.
The need to trust AI technologies grows stronger every day, as new smart technologies emerge. For instance, in the Agriculture Sector, AI is supporting and enhancing agricultural practices with automated robots that do more than traditional farming equipment.
Small Robot Company, a British-based engineering company, recently developed three robots—Tom, Dick, and Harry—to perform various tasks autonomously. Tom monitors soil and plants, Dick sprays fertilizers and chemicals, and Harry drills for various crops. When low on power, the robots return to their “kennel” to recharge.
Autonomous technology powered by AI relies on electronic equipment that meets the rules, regulations, and standards delineated by international and national committees. The rapid production of wireless products, mobile devices, and IoT and AI technologies has amplified the need for spectrum availability and spectrum planning. Wireless transceivers cannot distinguish between multiple transmissions on the same frequency, thus resulting in electromagnetic interference.
Rhein Tech Laboratories’ worldwide homologation services provides an unrivaled strategy for attaining regulatory product approval in over 100 countries. Our approach reduces the number of required emissions, immunity, and product safety tests by defining the minimum subset of regulatory standards for your target countries at the onset, reducing your time and cost to enter these marketplaces. Contact us today for questions.
Sources: https://www.nist.gov/sites/default/files/documents/2019/08/10/ai_standards_fedengagement_plan_9aug2019.pdf, https://iecetech.org/issue/2019-03/Robots-AI-and-big-data-pave-way-for-smart-farming
Comments