Private AI for enterprises in the UAE
Enclave42 gives UAE and GCC enterprises a way to run AI inference with confidential-computing protections for prompts and data during inference, rather than relying solely on ordinary public-AI processing — and to verify that themselves, rather than take our word for it.
What “private” means here
Every model in the Enclave42 catalogue runs as a confidential workload: processed inside an Intel TDX trust domain on an NVIDIA GPU. Enclave42 is not the trust anchor for that claim — Intel and NVIDIA are. You generate your own nonce, request the evidence, and verify it offline against their infrastructure, not ours.
Verify it, don’t just trust it
Attestation proves that the specific runtime instance serving you was genuinely executing inside an attested TDX trust domain on an attested GPU at the moment you requested evidence, and that the encryption key offered to you belongs to that instance. It does not audit what happens to your data once decrypted inside the enclave.
Access by plan
Explore, Build and Business plans unlock progressively wider model tiers, with usage billed separately through a prepaid wallet. Business, Sovereign and Enterprise customers can request private deployments, dedicated GPUs, and custom model hosting through a reviewed engagement.
UAE-resident infrastructure, where it’s contracted
UAE residency is not a property of every Enclave42 plan. On Explore, Build and Business, workloads run in attested enclaves whose physical location is not guaranteed. On the Sovereign plan, inference is executed on UAE-resident infrastructure, guaranteed in your contract — provisioned per customer, alongside a compliance pack and signed attestation reports you can archive.
Read the full breakdown of residency, sovereignty and compliance
Discuss Private AI for your organization
Tell us about your workload and we’ll walk you through what’s available on the standard catalogue versus what requires a Sovereign or reviewed engagement.