Implementing Amazon Bedrock Knowledge Bases in support of GDPR (right to be forgotten) requests | S02 EP43 | Lets Talk About Data
The General Data Protection Regulation (GDPR) right to be forgotten, also known as the right to erasure, gives individuals the right to request the deletion of their personally identifiable information (PII) data held by organizations. We will covering this aspect in out show with AWS Services like Bedrock
Tony Mullen
Amazon Employee
Published Nov 21, 2024
The show focused on implementing Amazon Bedrock knowledge bases to support GDPR requests, particularly the "right to be forgotten". We discussed the General Data Protection Regulation (GDPR), explaining it as a European regulation governing the protection of personal data. They emphasized that GDPR applies to any business handling EU citizen data, regardless of the company's location, with potential fines of up to 20 million euros or 4% of global revenue for non-compliance.
The discussion then moved to Amazon Bedrock, a fully managed service providing access to various foundation models for generative AI tasks. They explained how Bedrock knowledge bases can be used to implement Retrieval Augmented Generation (RAG), allowing organizations to combine their proprietary data with the capabilities of large language models. The hosts demonstrated how to create and use knowledge bases in Bedrock, showing how to upload data, query it, and handle data deletion requests in compliance with GDPR's right to be forgotten.
A live demo showcased the process of deleting user data from a knowledge base and verifying its removal from various components of the system, including the original data source, the vector database, and the foundation model's responses. The hosts emphasized the importance of ensuring data is removed from all relevant places when handling right to be forgotten requests in generative AI systems.
Highlights:
- GDPR applies globally to any business handling EU citizen data
- Amazon Bedrock provides managed access to foundation models for generative AI
- Knowledge bases in Bedrock support various data sources and file formats
- Implementing the right to be forgotten requires consideration of multiple data storage locations
- The demo showed how to verify data deletion across different system components
- Vector databases like OpenSearch Serverless, Pinecone, and Redis Enterprise Cloud can be used with Bedrock knowledge bases
- Supported file formats include PDF, Excel, CSV, HTML, DOC, text, and markdown
Check out the recording here:
Loading...
Tony Mullen - Senior RDS Specialist Solutions Architect @ AWS
Rajakumar Sampathkumar - Principal Technical Account Manager @ AWS
Yadukishore Tatavarthi - Senior Solutions Architect @ AWS
GPDR - https://gdpr.eu/tag/gdpr/
Any opinions in this post are those of the individual author and may not reflect the opinions of AWS.