Inside Anjuna’s Confidential Computing Approach to Data Protection

Why It Matters

By securing data during processing, Anjuna enables high‑risk sectors to adopt public‑cloud AI workloads without exposing sensitive information, reshaping cloud security economics.

Key Takeaways

• Seaglass runs unmodified apps inside hardware enclaves
• Confidential computing protects data while being processed
• AI Clean Rooms enable encrypted collaborative AI workloads
• $25M Series B2 funds NVIDIA Blackwell GPU integration
• 2026 projected as breakout year for confidential computing

Pulse Analysis

The traditional data‑protection playbook—encrypt at rest and in transit—leaves a critical blind spot when data is actively used. As AI accelerates attack cycles, breaches can occur the moment plaintext data resides in memory, exposing enterprises to insider threats and sophisticated malware. Confidential computing addresses this vulnerability by creating isolated processor enclaves that keep data encrypted throughout computation, effectively extending the perimeter to the hardware level and reducing reliance on a complex stack of perimeter defenses.

Anjuna Security’s flagship offering, Seaglass, abstracts the enclave complexity, allowing organizations to migrate legacy and modern applications into secure environments without rewriting code. This "digital battle armor" approach lowers adoption barriers and speeds cloud migration for workloads that were previously stuck on‑prem due to compliance or sovereignty concerns. The recent launch of AI Clean Rooms and the Northstar platform builds on this foundation, offering collaborative, privacy‑preserving AI model training that leverages NVIDIA’s Blackwell and Hopper GPUs. The $25 million Series B2 round, led by Insight Partners, fuels these integrations and expands Anjuna’s go‑to‑market capabilities.

The market implications are profound. Financial services, government, and defense sectors—where data leakage can trigger regulatory penalties—can now exploit public‑cloud scalability and AI innovation without compromising confidentiality. As AI agents become more autonomous, the need for hardware‑enforced guardrails intensifies, positioning confidential computing as a prerequisite for responsible AI deployment. Analysts anticipate a surge in enterprise spend on enclave‑based solutions, making 2026 a pivotal year for vendors that can deliver seamless, high‑performance, and compliant data‑in‑use protection.