01 Home 02 Company 03 Systems 04 Capabilities 05 Technology 06 Research 07 Investors 08 Careers 09 Newsroom 10 Contact

Responsibility

Autonomy you
can trust.

Autonomy for defence carries real responsibility. We treat safety, human command and accountability not as add-ons, but as design requirements — built in from the first line of code.

Our Commitments

Principles we design toward.

These aren't slogans. Each one shapes how we architect, build and assess our systems.

§ 01 / PRINCIPLES
01

Human In Command

A person is always in command. Machines carry out tasks; people decide whether they should happen, and remain responsible.

02

Human Oversight

Systems are designed so a person can supervise and intervene at any time — the leash is as long or short as the task requires.

03

Safety By Design

Clear operating limits and safe, predictable behaviour are built in from the start, not bolted on afterwards.

04

Lawful & Ethical Use

We build to be used lawfully and responsibly, within applicable rules and the direction of legitimate authorities.

05

Accountability

Decisions and runs are recorded so they can be understood and reviewed after the fact. Accountability by construction.

06

Honest Assessment

After every run we assess plainly — what went right, what went wrong, and why — and act on it.

07

Security

Sound security engineering as a default across the whole stack, from supply chain to runtime.

08

Restraint

We are deliberate about what we build and how we describe it. We would rather under-claim and over-deliver.

The Core Idea

People decide. Machines execute.

Autonomy is useful precisely because a system can carry out a task without being steered moment to moment. But carrying out a task and deciding whether it should happen are two different things — and the second always belongs to people.

Our systems are built to keep that line clear: bounded behaviour, a human able to supervise and stop them, and a full record afterwards. That is what makes autonomy something an institution can actually adopt with confidence.

GuardrailsBy Design
CommandA person is always in command
BoundsClear limits on what a system will do
InterveneSupervise and stop at any time
RecordEvery decision logged and reviewable
AssessHonest after-action review, every run

Talk To Us

Questions about our approach?

We're glad to discuss how we think about responsible autonomy.