Holistic AI, an AI governance platform for enterprises, has launched Holistic AI OSL, a free open-source library designed to help developers build fair and responsible AI systems. The Python library is available for download without any licensing requirements.

The library addresses five key technical risks in AI systems: Bias mitigation with over 35 metrics and 30 strategies, explainability features to reveal model decision-making, robustness testing for consistent performance, security measures for user privacy and protection against attacks, and efficacy testing to ensure balanced performance in real-world scenarios.

According to the company, the library aims to help organizations complete more AI projects successfully while avoiding risks, compliance issues, and bias. MAPFRE, a global insurer operating in nearly 40 countries, has already integrated the library into its AI innovation strategy.

Analyst QuickTake: The news follows the company's expanding platform access by making its AI governance solutions available on the Microsoft Azure Marketplace ( September 2024 ). Before this, the company introduced the Holistic AI Tracker Feed ( June 2024 ), which serves as a resource for updates on AI regulation and compliance.