Automated Audit of AI Algorithms

Many organizations are in the process of deploying AI applications into production. A major concern is how to make sure that the algorithms are trustworthy. Because AI models are new, there is no well-established framework yet to provide such guarantee to users and decision makers.

The EU made progress on this topic with the publication of the Assessment list for Trustworthy AI (ALTAI). The list contains seven categories of questions that help verify the quality of the algorithms.

1. human agency and oversight
2. technical robustness and safety
3. privacy and data governance
4. transparency
5. diversity, non-discrimination and fairness
6. environmental and societal well-being
7. accountability

With Chiron you can streamline the review of your AI application based on the ALTAI methodology. Thanks to pre-packaged templates, your team can create reproducible documentation that analyses how your model scores with respect to the seven categories of ALTAI. These documents will contain quantitative tests, developed by Yields.io, as well as qualitative questions that probe for the governance that has been put in place. Because the validation approach is data driven, it can as well be used for vendor models.