Standardization of Artificial Intelligence
When I mention that I work in AI, one of the first things that people ask me about is whether we will be unleashing killer robots that will be our downfall. While I’m fairly confident that is not going to happen, I do think that standardization and validation in artificial intelligence is an extremely important topic. I work with several amazing organizations to help make that a reality.
I was asked to join the ISO/IEC JTC 1/SC 42 committee on Artificial Intelligence upon its formation in November 2017 and in June 2018 I was elected Canada’s Head of Delegation for the Study Group on Computational Methods in AI. The committee is tasked with formulating Canada’s position on a variety of important areas to ensure our country remains a world leader in AI industry standards, computational methods, and policy to promote trustworthiness and confidence in AI among the general public. During the Fall 2018 Plenary in Sunnyvale, California, I was appointed as a liaison for ISO/IEC JTC 1/SC 37 Biometrics, where I will work on standards for AI as it applies to biometric data including facial recognition, fingerprint and retina scanning, genomics, and prevention of identity theft.
In the fall of 2018, I began working with Canada’s CIO Strategy Council on pan-Canadian standards for ethical design and use of automated decision engines. These standards are currently in development.
I also served on the Marketing Research and Intelligence Association’s Standards Committee from 2014-2018, and chaired the AI subcommittee. This committee worked as part of the self-governance of the Marketing Research industry to set Canada-wide standards for practitioners and companies within the industry. Throughout my tenure we successfully introduced new standards for many areas, including: ethical standards for data collected from children, youth, and vulnerable populations; transparency in data collection methods for political polling; standards for telephone interviewing and data collection using cell phones; and reporting standards for adverse medical events in the course of research on pharmaceutical products.
I speak frequently on the topics of standardization in AI, model validation, and privacy. Please see Speaking for more information.