I’ve spent my career in security, and I’ve never seen a moment like this. Every organization wants to use AI and machine learning to get better insights, modernize operations, and improve services. But the challenge is clear: how do you migrate sensitive data to the cloud, share it broadly, and still keep it secure and compliant?
That’s why I joined Dennis Duckworth and Joel Salinas from Cloudera for a conversation about data integrity, data protection, and why the old ways of thinking about security simply aren’t enough anymore.
Why Integrity Matters as Much as Security
At RSA this year, Bruce Schneier said something that stuck with me: “Integrity is a security property.” It’s not enough to lock data down. If data isn’t accurate, trustworthy, and fit for purpose, your AI models and analytics won’t deliver value.
Confidentiality and availability are well understood in IT. Integrity is different. It requires governance, quality, and trust in the data itself. Whether it’s training a model, informing executive decisions, or guiding customer interactions, integrity is what makes data usable.
Data Risks in the AI Era
AI brings enormous opportunity, but it also exposes new risks:
- More users: Natural-language tools mean more people are touching data, not just analysts.
- Unstructured data: Video, audio, documents, and IoT sensors are flowing into the mix.
- Dynamic pipelines: Data isn’t static at rest; it’s moving constantly across systems.
Traditional security models can’t keep up. Layers of firewalls and access controls add friction but don’t solve the problem. True protection requires embedding security directly into the data itself.
Data-Centric Security: Embedding Protection at Scale
At Protegrity, we call this data-centric security. Instead of surrounding data with barriers, you build protection into the data so it can move safely across environments. That means:
- Mobility at scale: Share and federate data across borders without losing control.
- Systems at scale: Apply protections consistently across on-prem, cloud, and multi-cloud.
- Cipher performance at scale: Use the right method—tokenization, anonymization, dynamic masking, synthetic data, or even homomorphic encryption—based on the workload.
There’s no silver bullet. You need a toolbox of protections tailored to how data is being used.
The Cloudera Perspective
Dennis put it well during our session: “You should be able to select the type of workload, select the capacity, run the job, and meet security requirements—wherever you want.”
That flexibility is what Cloudera offers. Their hybrid data platform allows customers to run workloads in the cloud, on-prem, or at the edge, all with consistent governance. Together, Cloudera and Protegrity give organizations both the freedom to innovate and the confidence to protect sensitive data at every step.
Real-World Impact
We’ve seen this approach at work across industries:
- Federal systems: Protecting $6.7 trillion in assets and securing a 24/7 instant payment system.
- Healthcare: Safeguarding three-quarters of the U.S. population’s health records while enabling research and outsourcing.
- Financial services: Securing 27 billion daily transactions across 100+ countries—without creating compliance silos.
In each case, the combination of Cloudera’s unified platform and Protegrity’s embedded protections allows organizations to scale securely and stay compliant, while still putting data to work.
Building the Secure AI Pipeline
AI raises the stakes even higher. To avoid what Joel calls the “POC junkyard” where 70–95% of projects stall, you need guardrails. That means:
- Curating secure datasets for training.
- Embedding privacy tools like anonymization and synthetic data generation.
- Using semantic controls to monitor how users and AI interact with data.
By embedding controls early in the pipeline, you avoid costly risks later. And by pairing Protegrity’s protections with Cloudera’s full AI lifecycle, you can move faster while staying secure.
See Us at Cloudera EVOLVE
If you’re serious about data security and AI, now is the time to act. Join us at Cloudera EVOLVE in New York City and London. There’s no cost to attend, and it’s your chance to see how Protegrity and Cloudera can help you build a secure, scalable foundation for the future of data and AI.
I’ll be there, and I’d welcome the chance to talk with you in person.