North Axiom Capital, home
Responsible AI

Controls built in, not bolted on.

How North Axiom governs the AI it builds, deploys and advises on.

Why it matters now

Regulation has arrived

The EU AI Act is now in force in stages, with substantial penalties. The UK takes a principles-based approach through existing regulators. Both expect demonstrable control.

€35m or 7%

Of global turnover, the maximum EU AI Act penalty for prohibited practices.

DLA Piper, EU AI Act obligations
2 August 2026

When the European Commission’s enforcement powers over general-purpose AI providers apply.

DLA Piper, EU AI Act obligations
51%

Of firms report an AI-related incident.

McKinsey, The State of AI in 2025

Eight controls

Applied to every system we build or operate.

AI governance

Named accountability, an approval route and a register of systems in use.

Human oversight

Explicit decision rights, approval thresholds and escalation paths.

Data

Lawful basis, minimisation, provenance and retention.

Security

Access control, secrets management, logging and monitoring.

Privacy

Personal data kept out of models and prompts unless there is a lawful basis.

Evaluation

Testing before release, and continuous monitoring after it.

Third parties

Model and vendor assessment, and contractual control of data use.

Regulatory alignment

EU AI Act, UK GDPR and sector rules. Fill in: confirm which regimes apply to North Axiom itself

The UK position

Principles-based rather than a single statute.

No comprehensive UK AI statute

AI is managed under existing rules on governance, resilience and data.

Financial services

Three quarters of UK financial services firms already use AI, and most have a named accountable person for it.

Discuss AI governance

Ask us how we would govern AI in your environment.