Our commitments

We build artificial intelligence for operations where mistakes are costly: public safety, emergencies, government data. That privilege comes with obligations. These are the rules we work by — and the standard you can measure us against.

01The state owns its data

Each institution's information is its own: it decides where it lives, how it is shared and when it is deleted. We do not use one client's data to train models for others, we do not sell it, and we do not keep it when the relationship ends. Our systems can run on client-owned, sovereign or air-gapped infrastructure when the project demands it.

02A human always decides

Our AI detects, prioritizes, translates and proposes — it does not dispatch patrols, impose sanctions or make decisions about people on its own. The operator keeps the context and the final word. Every product is designed to make its user more capable, not to replace their judgment.

03Everything is logged

Every search, every query and every access leaves a trace in an auditable log. Results cite their sources, so any figure can be defended before a supervisor, a council, an audit or a court. Technology that watches must itself be watchable.

04Privacy by design

We collect the minimum necessary, restrict access by role and apply the data protection rules in force in every deployment. Sensitive capabilities — such as recognizing individuals — are enabled only where the law allows it, under policies the client itself defines and can inspect.

05Security first

We encrypt information in transit, keep each client's environment separate and continuously test our systems. We report incidents transparently and fix them ahead of everything else.

06Clear limits on use

Our technology exists to protect people and make the state more effective — not for indiscriminate surveillance or the persecution of dissent. We document the intended uses of every product and reserve the right to walk away from projects that contradict them.

07Continuous improvement

Technology and the law evolve; so do our safeguards. We review our policies as rules and risks change, and we keep this page up to date so you always know what you can hold us to.