Tel Aviv, Israel and Atlanta, GA - January 31, 2023
Mona, the leading intelligent AI monitoring platform, announces a new feature that advances its monitoring capabilities with the introduction of a new feature for assessing AI fairness.
As many organizations began relying on AI in critical business functions, it has become important to identify and address algorithmic bias. The key to detecting such bias is understanding the behavior of machine learning models across different protected segments of the data within the context of the business function that the AI serves. Mona’s new AI fairness features make it simple to do that by providing a comprehensive view of the behavior of the AI system, as well automatically detecting any potential issues of bias.
With this new set of features, organizations are able to generate complete fairness reports (one-time or periodically). These reports include standard and custom (user configured) options and are applicable for both internal and external audit needs. Overall, fairness reports help organizations increase trust in their AI-driven applications.
“It is becoming increasingly important for companies that develop machine learning algorithms to determine whether biases are introduced when building datasets and training models,” said Giacomo Vianello, Principal Data Scientist at CAPE Analytics. “Mona has built tools and processes that are designed to ensure fairness, a critical step in building ethical AI.”