Assessing the Validity of a CAB’s Use of AI Under ISO/IEC 17025AI to Increase Lab Productivity in Environmental Testing
Oral Presentation
Prepared by J. Mensingh
Labtopia, P.O. Box 1228, Friendswood, TX, 77549, United States
Contact Information: [email protected]; 281-619-2600
ABSTRACT
As artificial intelligence (AI) tools become increasingly integrated into laboratory operations — from data processing and control chart analysis to automated reporting and instrument diagnostics — assessors will encounter these systems more frequently during ISO/IEC 17025 assessments. This session is designed specifically for assessors and focuses on how to confidently and consistently evaluate a conformity assessment body’s (CAB’s) use of AI while remaining grounded in the Standard and assessor scope.
Participants will learn:
Where AI is most likely to appear in laboratory processes, even when it isn’t labeled as such, and how to recognize risk without diving into technology.
Which ISO/IEC 17025 clauses are triggered by AI-assisted systems, with a focus on validity of results (Clauses 7.2 and 7.7) and the control of information systems (Clause 7.11).
A simple, assessor-centric framework built around two key questions: Does this impact the validity of results or decisions? And does the system process or manage data?
Plain-language questions and warning signs assessors can use in interviews and objective evidence collection to ascertain appropriate use, oversight, and validation.
Common missteps to avoid, including overreach into algorithm review or technical design, reinforcing that assessors are not expected to be AI experts but are expected to evaluate control, competence, and evidence.
