Responsible AI Dashboard Can Help Mitigate Risk

• Results of buggy or biased AI models can lead to serious consequences for organizations, including poor business decisions, reputational harm, and legal action.
• Microsoft established its Responsible AI principles in 2017 to help guide itself and customers on ethical use of AI.
• Microsoft’s Responsible AI Dashboard, in preview, can help organizations understand and mitigate risks and issues associated with their AI models.

 

Organizations increasingly rely on machine learning (ML, a popular form of artificial intelligence [AI]) to gain sophisticated insights into all aspects of their operations. However, the accuracy—and fairness—of ML inferencing depends on both the models and the underlying data used to train models. Organizations that use data not representative of target populations or groups may unknowingly derive biased AI results, which in turn can lead to consequences including legal action; some companies have begun to list AI bias as a risk on corporate disclosures such as 10-Ks, and/or have established corporate governance for AI initiatives. (See the sidebar “Understanding AI Bias.”)