Gartner® Report Securing Your AI Data Pipeline
In this report, Gartner offers key insights and recommendations for securing the data analytics pipeline for AI analytics and intelligence. Gartner breaks down the AI data pipeline into upstream, midstream and downstream, and provides guidance on where security controls are most effective.
- “High-priority data security controls on the AI data pipeline are metadata creation, differential privacy, data masking, tokenization and fully homomorphic encryption."
- "Applying controls upstream on the AI data pipeline is critical to mitigating issues downstream since few controls can be implemented there.“
- "In immature analytics environments where the AI data pipeline is not secured — especially in the downstream segment — chaos reigns and every data scientist could potentially fork some data away. Thus, data theft from the AI data pipeline is a major client concern."
- "Compliance with legislation and regulatory requirements such as privacy regulations, Basel II or the PCI DSS may require you to prevent exposure of confidential information to unauthorized entities and provide compliance reporting."
- "Leading organizations are embracing AI, but securing AI platforms and pipelines remains a key challenge because clients often fail to recognize the unique security concerns of their AI platforms."
- "Make your data easily consumable, via well-defined, protected methods, where you can enforce controls such as entitlement management, dynamic data masking or differential privacy on read and preventing toxic combinations of assets."
To read more Gartner recommendations on securing AI data pipelines, submit the form for complimentary access to this report.
Gartner, Securing Your AI Data Pipeline, Joerg Fritsch, August 2021
GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally, and is used herein with permission. All rights reserved.