Google Distributed Cloud: Running Gemini Where the Internet Can’t Go

There is a running joke in enterprise cloud sales: the most expensive words a government IT director can say are “we can’t use the cloud for that.” For classified workloads, air-gapped operational technology networks, and sovereignty-mandated environments, those words have been an ironclad ceiling for years. Google just started drilling through it.

Google Distributed Cloud (GDC) Air-Gapped is a fully managed cloud stack delivered as physical hardware that organizations operate entirely on-premises, with zero connectivity back to Google after delivery. Not a locked-down VPC. Not a compliance-certified region with a sovereignty badge slapped on the landing page. This is a rack of hardware that ships to your facility, Google sets it up, and then Google has no network path back in. Ever. The cryptographic air-gap is enforced post-deployment, not just promised in a terms of service.

What makes this genuinely new is what runs inside it: Gemini LLMs, Document AI, Speech-to-Text, Translation API, and Vision AI, all offline. Before GDC Air-Gapped, if you needed frontier AI on classified data, your options were build it yourself or don’t. Now there’s a third option that doesn’t require a team of cleared ML engineers and five years of runway.

Real Deployments, Not Just a Roadmap Item

The post-launch traction here is worth taking seriously. In July 2025, Google Public Sector was awarded a $200 million contract by the DoD’s Chief Digital and AI Office to accelerate AI and cloud capabilities across defense, and GDC Air-Gapped achieved DoD Impact Level 6 authorization, which covers secret classified data. By September, General Dynamics Information Technology ran the air-gapped appliance at the tactical edge during Exercise Mobility Guardian, demonstrating real-time language translation in a disconnected field environment. November brought a multi-million dollar NATO contract for sovereign cloud services using GDC Air-Gapped. Then in December, the Pentagon launched GenAI.mil, a generative AI platform for three million military and civilian personnel, with Google’s Gemini for Government as the first capability deployed on it.

Singapore, Australia Defence, and several EU sovereignty-mandate markets followed similar paths. This is not a press release product. It is a product with contracts, deployed hardware, and cleared facilities.

Why It Matters Beyond the Defense Market

The defense and intelligence angle is obvious. Yes, Google is chasing enormous multi-year government contracts. The impact extends further than Beltway deal rooms, though.

National data sovereignty laws in Germany, India, Saudi Arabia, and the UAE increasingly require cloud services to operate entirely within national borders with zero foreign provider access. Standard public cloud models, including so-called sovereign regions, still route management traffic through provider infrastructure somewhere abroad. GDC Air-Gapped is architecturally different: there is no management plane phoning home. For sovereign cloud providers in those markets, this is the difference between a compliance posture and an actual architecture guarantee.

Critical infrastructure operators face the same constraint from a different direction. Power grids, water treatment facilities, and financial market infrastructure run on isolated operational technology networks specifically because connectivity is a security liability. Cloud-like orchestration and AI have been unavailable in those environments by design. GDC scales from a single server to hundreds of racks, which means it fits both a compact edge OT deployment and a full sovereign data center.

The Competitive Picture

The competitive picture here is unusually clean, and it’s worth being direct about it.

AWS Outposts requires sustained AWS network connectivity for control-plane management. It is explicitly not air-gapped. Workloads can continue running locally during brief disconnections, but provisioning and management break. AWS knows this and has not tried to position Outposts as a true air-gap solution.

Microsoft has Azure Stack Hub, which supports a disconnected mode, and the newer Azure Local is making noise about offline AI capabilities with something called Foundry Local. In fully disconnected Azure Local environments, however, some native AI and ML services are unavailable, and certain configurations require periodic reconnection for license reporting. Microsoft is moving in this direction, but they are not there yet in the way GDC Air-Gapped is.

GDC Air-Gapped is currently the only hyperscaler offering that combines on-premises hardware delivery, a genuine post-deployment air-gap, DoD IL6 authorization, and Gemini running fully disconnected. That is not a marketing claim. It is a technical and compliance position that competitors have not matched.

Who Should Actually Care About This

The most direct ISV beneficiaries are those building for government and defense. If your product requires internet-connected cloud APIs today, you have been locked out of entire procurement categories. A defense ISV building a classified document intelligence platform can now deploy GDC at a government site and run Gemini for natural language search over classified military documents, with zero bytes leaving the facility. That is a product category that did not exist two years ago.

Sovereign cloud providers in markets with aggressive data localization mandates have an equally compelling case. Rather than building a Google-class AI platform from scratch, which is its own category of expensive and slow, they can use GDC Air-Gapped as the foundation for a national AI cloud service, offering Gemini, Vision AI, and Translation under a sovereignty guarantee that no Google infrastructure touches the data after delivery.

For ISVs in regulated commercial sectors, the implications are less obvious but still real. Healthcare networks with strict patient data controls, financial infrastructure with market isolation requirements, industrial automation vendors with OT network constraints: all represent TAM that was effectively zero before GDC Air-Gapped.

If you build software for government, defense, or regulated industries and you’ve been telling clients that AI on sensitive or even classified data isn’t practical, that conversation needs to be revisited.

Want to go deeper?