Google Cloud and Apple Partner on Confidential AI Infrastructure for Private Cloud Compute
Key Takeaways
- ▸Google Cloud and Apple have partnered to power Apple's Private Cloud Compute on Google Cloud infrastructure with custom-built security and privacy technologies
- ▸The platform leverages Google's Confidential Computing portfolio and Titanium architecture with custom Titan chips to protect data in use during AI inference via hardware-based TEEs
- ▸Multi-layer security approach combines hardware isolation (Intel TDX, NVIDIA Confidential Computing), Titan hardware root of trust, and open-source transparency mechanisms for independent verification
Summary
Google Cloud has announced a strategic collaboration with Apple to power Apple's expanded Private Cloud Compute (PCC) systems on Google Cloud infrastructure, announced at WWDC 2026. The partnership leverages Google's Confidential Computing portfolio and proprietary Titanium security architecture, which features Google's custom-designed Titan chip, to ensure data protection throughout its lifecycle—at rest, in transit, and critically during processing within hardware-based Trusted Execution Environments (TEEs).
The serving platform built jointly by Apple and Google is designed to meet rigorous security, confidentiality, and transparency requirements, with multiple layers of protection including hardware isolation via Intel TDX and NVIDIA Confidential Computing, Titan chips for hardware root of trust, and open-source host stack components enabling independent security verification. This multi-layered approach ensures that PCC infrastructure provides enforceable protections with no privileged runtime access and complete transparency into the system's security properties.
The collaboration represents a significant milestone in building secure cloud infrastructure for AI workloads, incorporating technologies and standards from Apple, Google Cloud, Intel, and NVIDIA. By ensuring that every layer of the stack—both hardware and software—contributes to verifiable security, the partnership addresses growing demands from enterprises running sensitive AI inference workloads while maintaining strong privacy guarantees.
- Partnership represents a major advancement in confidential AI infrastructure, addressing enterprise needs for secure, private AI workload execution with verifiable security guarantees



