By Paolo Campoli
Founder of Campoli Consulting
As someone who has spent decades in the service provider industry, I’ve seen firsthand how technology can transform businesses and how quickly new opportunities can become new risks. AI is the latest frontier, and while it promises incredible value, it also brings serious concerns about security and compliance. That’s why the partnership between Tumeryk and DataKrypto is so important and why I’m excited to see how this strategic integration will shape the future of secure, responsible AI.
No enterprise can compete without AI. But feeding an LLM with your confidential data—or, even worse, your customers’ data—exposes your company to significant risk and may even violate NDAs and compliance requirements. Once the LLM is trained with that data, malicious actors can use extraction and reverse engineering techniques to recover sensitive information.
While self-hosting an LLM might seem like a viable option, it’s not practical for most mid-sized enterprises. And in a world of Agentic AI, no model—self-hosted or not—will remain isolated. The core problem holding back AI adoption for Small and Medium-sized Enterprises (SMEs) and derailing sovereign AI initiatives for many governments is this: how can we feed the beast with encrypted information, compute on encrypted data using an encrypted LLM, and return encrypted results without ever exposing that data? And critically, how can we ensure the LLM is trustworthy before encryption, remains trustworthy after encryption, and that data flows securely across the entire AI pipeline?
That’s where the unique partnership between DataKrypto and Tumeryk comes into play. The combination of DataKrypto’s FHEnom for AI™, providing continuous AI encryption, and Tumeryk’s real-time encrypted model monitoring, creates a first-of-its-kind Secure AI framework. Campoli Consulting further strengthens this framework, helping organizations tailor and implement these advanced security solutions to address their specific challenges and regulatory needs.
Traditional security measures often focus only on what comes out of AI models, leaving the underlying data flows exposed. With regulations like General Data Protection Regulation (GDPR), Health Insurance Portability and Accountability Act (HIPAA), California Consumer Privacy Act (CCPA), and the upcoming EU AI Act, this simply isn’t enough. Organizations need a way to protect sensitive data at every stage, from the moment it enters the system to the final output.
Tumeryk and DataKrypto have partnered to create the industry’s first Encrypted Guardrails for Operational Security. FHEnom for AI keeps data encrypted throughout the entire process from vector embeddings, model weights to prompt payloads. Tumeryk’s Guardrails then inspect these encrypted context windows in real time, blocking any toxic or non-compliant prompts before they reach the model. The result? A seamless, automated compliance framework that maps to standards like the NIST AI Risk Management Framework (AI RMF), ISO 42001, Payment Card Industry Data Security Standard (PCI DSS), and System and Organization Controls 2 (SOC 2).
For Managed Service Providers (MSPs), this is a unique opportunity. As they transition to a Model-as-a-Service delivery model, MSPs can now offer not just performance, but end-to-end PII protection, all while helping customers meet regulatory and internal policy requirements. This is also a breakthrough moment for governments and institutions looking to realize sovereign AI that protects home-grown innovation, preserves citizen privacy, and keeps national IP secure.
John D. Boyle puts it best in his recent article:
“In an age of accelerating model adoption, regulatory pressure and constant AI drift (not to mention AI Shiny Object Syndrome), their (DataKrypto and Tumeryk) collaboration marks a turning point for how enterprises approach and operationalize AI security.” (From Security Boulevard, June 24, 2025)
Luigi Caramico, Founder and CTO of DataKrypto also adds:
“Our customers are clear. They want to harness the power of generative and agentic AI—but not at the cost of data security or compliance. With FHEnom for AI, we give enterprises a way to move fast and stay secure. They can now run powerful AI models on encrypted data without exposing sensitive information at any point. It’s a game-changer for organizations that need to protect user privacy, safeguard intellectual property, and maintain trust while scaling AI across their business.”
I’ve known Luigi Caramico for decades since our early days working with Italian service providers. Luigi was always relentlessly focused on improving how telcos manage big data and security. Our shared history gives me a unique appreciation for DataKrypto’s vision and the impact of this partnership with Tumeryk.
Campoli Consulting is proud to partner with Tumeryk and DataKrypto to bring this solution to MSPs in Europe and worldwide. With zero-trust AI, built-in compliance, and plug-and-play integration, MSPs can now offer cutting-edge, secure AI services to their clients without the usual headaches.
Encrypted Guardrails for Operational Security is available now as a Managed SaaS or self-hosted solution. For MSPs and enterprises ready to lead in secure AI adoption, this is your moment and we look forward to supporting you as we embrace this next evolution in AI protection and innovation.
About the Author
Paolo Campoli is Founder of Campoli Consulting and a strategic advisor to MSPs and critical infrastructure operators. He helps organizations navigate the complexities of secure digital transformation and AI adoption, drawing on decades of experience in the Italian and global service provider industry. Connect with Paolo on LinkedIn.