Legal AI company Cicero has selected Australe AI, citing risk management and greater control over sensitive data as key drivers behind the decision.

The move will see Cicero use AI infrastructure hosted and operated in Australia, enabling its data and AI workloads to remain within Australian jurisdiction and giving the company greater control over where sensitive information is processed.

For Cicero, whose platform handles sensitive, proprietary and regulated information, the decision reduces unnecessary exposure of customer data to offshore public AI platforms while limiting its dependence on offshore hyperscale AI providers.

Cicero’s CEO, Joseph Rayment, said: “For us, sovereignty isn't an abstract technology issue. We're working with highly sensitive legal information, so knowing where that data is processed and which jurisdiction it sits within matters. Australe allows us to keep that control without compromising on the AI capability we can offer our customers”.

The move also gives Cicero access to production-ready AI infrastructure without the time and cost of designing, deploying and managing its own environment. By removing the complexity of GPU infrastructure, model deployment and ongoing platform management, Australe enables Cicero to focus on developing and delivering its legal AI platform.

For Australe, Cicero’s decision provides a significant demonstration of the role sovereign AI can play in supporting real-world, business-critical applications.

Australe’s Chief Strategy Officer, Steven Duggan said “Cicero is putting AI to work in an environment where performance really matters. Our role is to make sure the infrastructure underneath it is ready for that—reliable at scale, built for production and flexible enough to evolve as Cicero does”.

The selection also gives Cicero greater flexibility as the AI landscape continues to change. New models and capabilities can be adopted through Australe without Cicero having to fundamentally redesign its application each time the underlying technology evolves.

The move comes as Australian organisations face growing questions around the jurisdiction, governance and infrastructure behind generative AI, particularly where sensitive or regulated information is involved. For companies building AI directly into their products, the decisions being made now are increasingly not just about which model to use, but what sits behind it.