Applied Artificial Intelligence and Machine Learning Test and Evaluation Program for Consumer Products
This report proposes a four-step test and evaluation program for artificial intelligence and machine learning technologies embedded in consumer products. Created by the U.S. Consumer Product Safety Commission's Chief Technologist, the program outlines methodologies to screen for AI/ML components, assess their implications, analyze their contributions to safety hazards, and monitor their evolution over time. The document provides recommendations for establishing testing protocols through a phased approach and collaborating with external standards bodies.
Summary and content fields are drawn from the document text and are pending editorial review.
Who issued it
Who it governs
This report is for the Commission and stakeholders interested in how product safety regulators approach testing and evaluating artificial intelligence and machine learning in consumer products.
What the document contains
The document covers a proposed four-step test and evaluation framework for identifying and analyzing artificial intelligence and machine learning components in consumer products. It outlines a three-phase implementation approach ranging from individual feature testing to comprehensive environmental evaluation, and details specific screening metrics for data, compute, and connectivity.
The document notes that it has not been reviewed or accepted by the Commission, and it does not establish final mandatory regulatory standards or legally binding rules for manufacturers.
- The CPSC Chief Technologist proposed a four-step test and evaluation program to assess AI/ML technologies in consumer products.
- The program follows four core steps: screen for AI/ML components, assess capabilities, analyze contributing factors to safety hazards, and monitor/measure evolution.
- Developing the testing capability is structured across three progressive phases covering specific functions, integrated systems, and conditional/environmental impacts.
- The document emphasizes collaboration with external organizations like NIST, UL, and ACT-IAC to develop standards and share best practices.
· EXECUTIVE SUMMARY · INTRODUCTION · BACKGROUND · PROPOSED TE PROGRAM - STEP-1 Search: Identify AI/ML Components - STEP-2 Assess: Implication of AI/ML Capabilities - STEP-3 Analyze: Impact of AI/ML Concerns - STEP-4 Monitor/Measure: Iteration of AI/ML Evolution · RECOMMENDATIONS
How it scores
Each score is backed by a quote verified against the document text. Editorial review has not yet confirmed them.
Accountability
2/5Clear responsibility for AI outcomes
To evaluate a product where AI/ML are present, a Subject Matter Expert (SME) would determine what capabilities exist, and document them into functional categories for further analysis.
PROPOSED TE PROGRAM
screening score · 2026-10-23
Fairness
0/5Non-discrimination and equitable treatment
screening score · 2026-10-23
Human Dignity
0/5Respect for the inherent worth of persons
screening score · 2026-10-23
Human Oversight
0/5Humans remain in control of consequential decisions
screening score · 2026-10-23
Privacy
0/5Protection of personal data and private life
screening score · 2026-10-23
Transparency
2/5Openness about how AI systems work and are used
The operation of consumer products has become less transparent with the introduction of AI/ML technologies.
BACKGROUND
screening score · 2026-10-23
Enforcement
1/5Consequences and compliance mechanisms with real force
Issuing and enforcing mandatory standards to address consumer product safety hazards due to products with AI/ML technologies.
INTRODUCTION
screening score · 2026-10-23
Impact Assessment
5/5Required assessment of an AI system's risks/impacts before or during use
Ultimately the four‐step process to screen, assess, analyze, monitor and measure is the core of the applied AI/ML TE program.
PROPOSED TE PROGRAM
screening score · 2026-10-23