Are you working to draft policy for your organization around implementing Trustworthy AI?
The following resources can be used to help you ensure your organization develops a framework that is rooted in ethical and trustworthy AI:
Many of the points in these resources underscore the importance of transparency, explainability, bias mitigation, and risk management-- all of which can be managed and addressed with watsonx.governance. As many organizations, as well as state and local governments are developing their AI policies and regulations, these are important points to keep in mind to implement a framework that prepares your organization for today's emerging technology and advancements to come.
What do I mean by this?
While we may not know which regulations are around the corner implementing an organizational standard and method by which your organization is able to ensure a new model meets and exceeds a regulation is a best practice. Similarly, we may not know what new metadata related to a model is yet to come, but having a system to manage model metadata (such as training data, bias/ quality/ drift metrics, and change requests) is a great way to set your organization up for success. Another example is to implement a framework to request a new model, along with an ethics board to review, assess risk, and approve new use cases.
We are in a booming period for incorporating AI into businesses at a rate faster than ever. It is essential to ensure we are implementing frameworks to govern these models and ensure they are trustworthy.
Lindsey Sample
lindseysample@ibm.com
https://www.linkedin.com/in/lindseysample/
#watsonx.governance