S. 1633: Testing and Evaluation Systems for Trusted Artificial Intelligence Act of 2025

This bill, titled the Testing and Evaluation Systems for Trusted Artificial Intelligence Act of 2025, proposes the establishment of a pilot program by the Director of the National Institute of Standards and Technology (NIST). The primary aim of this program is to create and implement measurement standards for evaluating artificial intelligence (AI) systems used by Federal agencies. Here are the key components of the bill:

Definitions

The bill defines several terms critical to its operations, including:

• Artificial intelligence system: This refers to systems as defined by prior legislation regarding AI.
• Covered foreign country: Includes specific nations deemed as potential adversaries, such as China and Russia.
• Testbed: A facility or mechanism for conducting rigorous testing of technologies to assess their functionality and performance.

Pilot Program Implementation

The Director of NIST, in cooperation with the Secretary of Energy and a newly formed Artificial Intelligence Testing Working Group, will oversee a pilot program. The program will:

• Utilize testbeds to evaluate how well AI systems work in practice.
• Enable an iterative process where measurement science is refined and standards developed.
• Ensure that the results are reviewed by stakeholders from the government, private sector, and academic institutions.

Memorandum of Understanding

Within 180 days of the bill's enactment, there will be a coordination agreement between the Secretary of Commerce and the Secretary of Energy. This agreement will provide access to necessary resources and personnel from the Department of Energy to support the pilot program.

Working Group Formation

Within 90 days of signing the coordination agreement, the Secretary of Commerce will form the Artificial Intelligence Testing Working Group. This group, which will not exceed ten members, will include representatives from:

• The Department of Commerce
• The Department of Energy
• NIST
• The private sector
• Academic institutions

Citizens from the designated covered foreign countries will not be permitted to serve on this group.

Measurement Standards Strategy

Within one year of its establishment, the Working Group must develop a strategic plan for creating measurement standards for AI evaluation. This plan should:

• Identify necessary measurement standards related to reliability, performance, security, and more.
• Include a blueprint for developing these standards.
• Identify specific AI applications for which the standards will first apply.

Development of Testbeds

Within two years of the bill's enactment, NIST, with input from the Working Group, will develop testbeds to demonstrate the measurement standards for AI systems. The Director of NIST is also authorized to hire experts as needed from various sectors to assist in this development.

Reporting Requirements

After the completion of the initial testbed demonstrations, NIST must report to Congress within 180 days. This report will summarize findings, suggest improvements to the pilot program, and recommend any necessary legislative or administrative actions.

Relevant Companies

• GOOGL (Alphabet Inc.): As a major player in AI technology, any new standards may impact its AI development protocols and competitive position.
• AAPL (Apple Inc.): Similar to Google, Apple is heavily invested in AI for its products and services, potentially altering how it develops and evaluates its AI technologies.
• MSFT (Microsoft Corporation): With a significant presence in AI through various products, Microsoft's AI evaluation strategies could be influenced by new standards established under this bill.