Elite members get real-time data, technical charting, alerts, and no ads - starting at $24.96/mo.
Try Elite for Free

The Finviz API:   Higher Limits, Clearer Data, and More Ways to Connect

Get Started Now

Ondas and Palantir Launch Regna AI to Develop Sovereign Artificial Intelligence for Defence Operations

By Fiona Craig | October 12, 2026, 8:32 AM

Ondas Inc. (NASDAQ:ONDS) and Palantir Technologies Inc. (NASDAQ:PLTR) have announced the establishment of Regna AI, a privately held company focused on developing mission-specific artificial intelligence systems for defence, security and other operational environments.

The new business will initially be wholly owned by Ondas and will use Palantir’s technology platform to enable customers to train, customise and deploy AI models within infrastructure under their own control.

According to the companies, Regna AI is designed to address growing demand for sovereign AI capabilities, allowing government and commercial customers to retain ownership and control of sensitive mission data and the models developed using that information.

Ondas will become Regna AI’s first customer, deploying specialised AI models to support autonomous mission orchestration at the edge, where computing systems operate directly alongside equipment and sensors rather than relying exclusively on centralised cloud infrastructure.

Palantir will contribute model-training infrastructure and forward-deployed engineering support, providing technical resources for the development and implementation of customer-specific AI systems.

The initiative builds on an existing collaboration between the two companies involving SkyWeaver C4I, an open-architecture command, control, communications, computers and intelligence programme.

SkyWeaver C4I is designed to integrate surveillance capabilities across multiple operational environments, ranging from stratospheric platforms to tactical systems operating closer to the ground.

Regna AI will extend this approach by developing AI models intended to operate across a range of hardware platforms, including unmanned aerial vehicles, stratospheric balloons, ground vehicles, surface vessels and command-and-control systems.

The companies said the technology is being designed for deployment in environments where communications may be disrupted, degraded, intermittent or otherwise restricted.

By allowing AI models to operate on local hardware, Regna AI aims to support mission execution and decision-making without requiring continuous access to external networks or centralised computing resources.

Eric Brock, Chairman and Chief Executive Officer of Ondas, said maintaining control over AI infrastructure and operational data was an important consideration for government customers.

“No government should have to build its AI strategy on models or data infrastructure it does not control,” Brock said.

Alex Karp, Co-Founder and Chief Executive Officer of Palantir, highlighted the potential advantages of developing AI systems trained specifically for operational requirements.

“Sovereign AI, trained on mission knowledge and tradecraft, outperforms frontier models while removing dependence on them,” Karp stated.

The companies intend to expand Regna AI’s applications beyond defence into security, critical infrastructure and industrial markets, where organisations may require specialised AI systems operating within controlled environments.

Ondas and Palantir said they will collaborate on scaling the technology platform and supporting initial customer deployments over the coming months.

Financial terms associated with the formation of Regna AI, including investment commitments and potential revenue contributions, were not disclosed.

The announcement establishes a new commercial structure for the companies’ existing technology collaboration, with Ondas providing the initial customer relationship and Palantir supplying the underlying AI development infrastructure.

Ondas Holdings stock price

Palantir Technologies stock price

Latest News

1 hour
1 hour
1 hour
5 hours
7 hours
Oct-11
Oct-10
Oct-10
Oct-09
Oct-09
Oct-09
Oct-09
Oct-09
Oct-09
Oct-09