i2 develops software that helps people make sense of complex information and make decisions in high-stakes environments, including law enforcement, defence, intelligence, and national security.
As AI becomes part of that work, our position is simple: AI must be trusted to be useful.
In these environments, trust is not optional. Decisions carry legal, operational, and societal weight. That means AI must be transparent, controlled, and grounded in evidence.
Our teams include experienced intelligence practitioners with first-hand understanding of operational realities. This shapes how we design and govern AI, ensuring it reflects the caution, scrutiny, and evidential standards expected in high-trust environments and supports users in making informed, defensible decisions.
Our Approach to AI
Human-assisted by design
AI supports analysis. It does not replace human judgement. Users interpret outputs, apply context, and make the decisions.
Transparent by default
We are clear about where and how AI is used, what it is designed to do, and where its limits are. Where AI influences outcomes, we prioritise approaches that make those outcomes understandable.
Built for control and validation
AI is embedded in workflows that support review, challenge, and verification. Wherever possible, it runs in customer-controlled environments so organisations stay in control of their data, models, and governance.
Our Commitments
Protecting customer data
We do not use customer data to train or fine-tune models, and we do not share it without explicit agreement.
Human oversight
AI outputs are always subject to review. Responsibility for decisions remains with the user, supported by workflows that enable validation and challenge.
Clear boundaries on AI use
i2 solutions support decision-making across complex operational environments. They are designed to inform and support human judgement, not to make autonomous, legal, or operational decisions.
Built on Trusted Frameworks
Our AI approach is supported by structured, auditable governance designed for mission-critical environments.
Lifecycle governance
Clear processes across design, development, deployment, and ongoing oversight, aligned to recognised standards such as ISO/IEC 42001 to ensure accountability across the AI lifecycle
Alignment with emerging regulation
Continuous review against evolving global standards, including the EU AI Act, with a focus on transparency, human oversight, and risk management
Evidence-led assurance
AI features are assessed before release to ensure appropriate controls, traceability, and transparency, supported by documented governance processes and auditability
How AI Works in i2 Products
For over 35 years, i2 products have used advanced analytical techniques, including deterministic methods such as entity resolution and network analysis, to support complex investigations. These approaches are proven, repeatable, and trusted in operational use.
Today, we are extending this foundation with machine learning and natural language capabilities, developed in close collaboration with our customers.
The principles remain consistent:
AI enhances human analysis
Outputs can be understood and scrutinised
Control stays with the user
Looking ahead
We will continue to evolve our use of AI in partnership with our customers, focusing on usability, explainability, and control.
Our position will not change:
AI supports informed human decision-making. It does not replace it.
