Privacera, the AI and data security governance company founded by the creators of Apache Ranger™, has officially announced the launch of new access control and fine-grained data filtering functionality for Vector DB/RAG. According to certain reports, the stated functionality will join the company’s Privacera AI Governance (PAIG) platform, which is the industry’s first comprehensive generative AI governance solution. Talk about the stated platform for a second, PAIG basically leverages GenAI technologies to secure the entire AI application lifecycle, right from discovering sensitive data, RAG, and user interactions feeding into AI-powered models to model outputs and continuous monitoring of AI governance through comprehensive audit trails. Now, as for how the development in question will make that mechanism even better moving forward, the answer begins from a seamless integration with multiple data sources. Here, users will enjoy the facility to merge data from varied platforms like Confluence, SharePoint, Databases, and support tickets into VectorDB. Such a facility, on its part, will go a long distance to ensure that original access policies of these sources are accurately reflected for users and groups. Next up, we must get into the promise of advanced classification-based filtering, a promise which empowers users to implement robust security and compliance policies through classification and tagging of data segments in VectorDB. An example of the same would be how access to finance-related data in VectorDB can be restricted exclusively to members of the finance team. Beyond that, such a facility can also come in handy for keeping embeddings tagged as “INTERNAL” from being supplied as context to the LLM whenever contractors or external users query the GenAI applications.
“In generative AI, Retrieval-Augmented Generation (RAG) systems operate by sourcing contextual information from a VectorDB, aggregating data from diverse origins such as Confluence Wiki pages, SharePoint, Databases, and support tickets, and other operational systems. These sources inherently possess their own access controls, so it’s crucial that the VectorDB inherits those and then maintains and enforces equivalent security measures when utilizing this data in generative AI applications,” said Don Bosco Durai, co-founder and CTO at Privacera. “PAIG makes it easy to maintain distinct access controls aligned with the original source permissions.”
Rounding up highlights for us are these fine-grained authorization protocols that enable users to employ dynamic metadata filtering to tailor access rights, and therefore, guarantee real-time compliance and heightened security. To again contextualize the picture with an example, we can acknowledge how users will have the opportunity to enforce GDPR and CCPA by filtering customer data based on geographic location or individual consent.
Founded in 2016, Privacera has risen up on the back of its SaaS-based unified data and AI security platform, which fetches for you cutting-edge data privacy, security, and governance capabilities. The company’s excellence in what it does can be understood once you consider it is currently serving clients from various different sectors, including finance, insurance, life sciences, retail, media, consumer, and government etc. Having achieved the AWS Data and Analytics Competency Status, Privacera further collaborates with leading data sources, such as AWS, Snowflake, Databricks, Azure, and Google. In case that didn’t sound like impressive enough, then we must mention how the company has also been recognized as a leader in the 2023 GigaOm Radar for Data Governance, a finalist for 2022 CISO Choice Awards, as well as “Sample Vendor” in the Gartner® Hype Cycle™ for Data Security.