According to a new press release, Privacera, the pioneering AI and data security governance company founded by the minds behind Apache Ranger, has unveiled new features enhancing access control and fine-grained data filtering for Vector DB/RAG within Privacera AI Governance (PAIG). Don Bosco Durai, Privacera’s co-founder and CTO, emphasized the importance of maintaining original source permissions when utilizing data from diverse origins like Confluence Wiki pages and databases in generative AI applications. PAIG ensures seamless alignment and enforcement of security measures, vital for robust user- and group-level policy enforcement within Vector DB.

The latest enhancements to PAIG aim to fortify organizations’ ability to secure sensitive data and mitigate risks associated with AI applications. By establishing and administering access control policies and facilitating fine-grained data control and filtering, PAIG empowers organizations to accelerate their generative AI product strategies while safeguarding data integrity. PAIG oversees the entire AI application lifecycle, from securing fine-tuning data and user interactions to monitoring AI governance through comprehensive audit trails.

PAIG’s upgraded features include seamless integration with multiple data sources such as Confluence, SharePoint, and databases, ensuring accurate reflection of original access policies. Advanced classification-based filtering enables robust security and compliance policies, while fine-grained authorization protocols allow dynamic metadata filtering to tailor access rights in real time, ensuring compliance with regulations like GDPR and CCPA. These enhancements streamline the process of managing contextual data, empowering users to enforce robust security measures within Vector DB effortlessly.