{{item.title}}
Different businesses and industries have diverse biases in their AI models. Customize your approach to finding bias by choosing the relevant metrics and thresholds for your needs, risk tolerance and governance policies.
Continually improve governance processes by monitoring AI model bias thresholds and potential data disparity issues, which could cause discrimination based on race, age, gender and other characteristics.
Before making changes to your AI models, you can get specialized recommendations for different mitigation strategies, simulate their potential impact and compare the results to fairness metrics.
PwC’s web application, Bias Analyzer, created by our Technology Impact Office, is recognized in both Fast Company’s first-ever Next Big Things in Tech 2021 class of honorees, and CogX Awards 2022 for Outstanding Achievements & Research Contributions in Algorithmic Bias Mitigation.
Mitra Best
Technology Impact Leader, PwC US