Direct answer: Ondas and Palantir announced on 12 October 2026 that they have formed Regna AI, a standalone private company building mission-specialized AI models that governments and defence organisations can run inside their own secure environments. Palantir provides the technology backbone, while Ondas — a maker of autonomous drones and defence systems — is the first customer.
What happened
The two companies announced the formation of Regna AI on Monday via an official press release. Regna AI is a standalone private company, described as “powered by Palantir” and “bootstrapped by Ondas,” and is initially wholly owned by Ondas. The companies say it is expected to scale with institutional partners over time.
The concept is sovereign AI: mission-specialized, fine-tuned models deployed within the customer’s own environment, so mission data and infrastructure stay under customer control. That reduces dependence on externally controlled models and avoids the cost of building frontier models from scratch. Regna AI will use Palantir’s AIP platform and open-weight models to deliver those capabilities.
Ondas will serve as Regna AI’s first customer, deploying fine-tuned models for autonomous mission orchestration at the edge. The target platforms read like a defence catalogue: drones, stratospheric balloons, ground vehicles, surface vessels, and command-and-control nodes — designed to work in denied, degraded, intermittent, or limited connectivity conditions. The models will help operators assess information and propose mission-plan updates for human review and authorisation.
Palantir is contributing model-training infrastructure and engineering support, building on the companies’ existing SkyWeaver C4I collaboration. Initial customer deployments are expected over the coming months. The launch targets defence first, with security, critical infrastructure, and industrial markets to follow.
Ondas chairman and CEO Eric Brock put the pitch bluntly: “No government should have to build its AI strategy on models or data infrastructure it does not control.”
What it means for UK small businesses
This is a defence announcement, not a product you can buy. But the idea underneath it — sovereign AI — is drifting steadily toward ordinary business, and UK firms should understand it now.
The questions Regna AI answers for governments are the same questions GDPR already forces on UK businesses: where is our data processed, who controls the model, and what happens if the vendor changes the rules? A small firm choosing an AI supplier in 2026 is, in a small way, making the same sovereignty calculation — just with customer records instead of mission data.
Practically, expect “sovereign AI” to become a sales pitch aimed at you within a year or two: vendors offering models that run on your infrastructure, in your region, under your control. When they do, you’ll want to know what you’re buying. Regna AI is the defence-grade, Palantir-powered version of that shift — and a useful reference point for separating genuine control from marketing.
If data handling already keeps you up at night, our GDPR and AI guide for UK small businesses covers the compliance side of the same problem.
Honest limits / what to watch
- No technical detail. The announcement contains no benchmarks, no model names, and no architecture specifics. “Mission-specialized fine-tuned models” is a claim, not evidence.
- “Sovereign” still runs on Palantir. Customer-controlled infrastructure is the promise, but the platform underneath is Palantir’s AIP. Sovereignty here means control over data and deployment, not independence from the vendor.
- Defence first, everyone else later. Initial deployments are defence-focused. The promised expansion into security, critical infrastructure, and industrial markets has no timeline.
- One named customer. Ondas is both co-founder and first customer. Until independent governments or institutions sign on, this is a partnership with a roadmap, not a proven platform.
Sources: Ondas press release, Morningstar / ACCESS Newswire, Crypto Briefing
FAQs
What is Regna AI?
Regna AI is a standalone private company formed by Ondas and Palantir on 12 October 2026. It develops mission-specialized, fine-tuned AI models that governments, defence organisations, and institutions can operate within their own secure environments.
Who owns Regna AI?
Regna AI is initially wholly owned by Ondas and described as “powered by Palantir” and “bootstrapped by Ondas.” The companies say it is expected to scale with institutional partners over time.
What does sovereign AI mean?
Sovereign AI means AI systems whose models, data, and infrastructure are controlled by the customer rather than a foreign vendor or cloud provider. For governments, it means mission data never has to leave their own secure environment.
How is Regna AI different from using a commercial AI model?
Commercial models run on the provider’s infrastructure with the provider’s rules. Regna AI’s models are designed to be fine-tuned for specific missions and deployed on customer-controlled hardware — including drones and command nodes in low-connectivity environments — with human review kept in the loop.
When will Regna AI deploy?
The companies say they expect to support initial customer deployments over the coming months. Ondas will be the first customer, deploying fine-tuned models for autonomous mission orchestration at the edge.
Why should UK small businesses care about a defence AI launch?
Sovereign AI is a wider trend, not just a military one. The same questions — where is our data processed, who controls the model, can it run on our infrastructure — are reaching ordinary businesses through GDPR and vendor scrutiny. Regna AI is the defence-grade version of a shift every UK firm will feel.
