Methodology

Disciplined Human–AI collaboration

Strategic Horizon treats AI as an institutional research and review capability while keeping evidence, inference, judgement and decision authority distinct.

Evidence before assertion

Public analysis distinguishes what is directly supported by sources from what is calculated, inferred or assessed. Known source gaps are not silently filled by assumption.

Claims are bounded

Important statements are framed with the time, geography and institutional scope needed to keep them independently quotable without changing their meaning.

Adversarial review

Review is proportionate to significance, risk and urgency. Material claims may be challenged through red-team, legal, risk or strategic review according to their substance.

Falsifiability by design

Where useful, an assessment identifies what evidence would materially weaken, change or refute it. The objective is not to protect a house view, but to expose it to evidence.

Human–AI institutional model

Specialised AI agents support retrieval, synthesis, challenge and expert review inside defined institutional roles. Human leadership retains judgement, accountability, external commitments and reserved decisions.

Public methodology, not internal prompts. This page publishes research standards and institutional principles. Internal instructions, credentials, security details and private operating data remain outside the public layer.