Our AI commitments
AI, in its place
AI assists processing. It does not replace evidence or human judgement.
These commitments apply to our consulting and accompaniment projects.
Evidence and judgement — human-led
- Sources: reports, articles and expert interviews. We consult original sources directly; a model’s response is not evidence.
- Weak signals and emerging tensions: identified through our research, not accepted on a model’s authority.
- Start-ups, pioneering projects and partners: drawn from our curated industry and innovation lists. AI may suggest additions; we verify each one before including it.
- Strategy, trade-offs, facilitation and collective intelligence: human responsibilities, never delegated to a model.
Processing — where AI helps
- Coding, extracting data from reports and structuring databases: technical work we might previously have scripted in Python. Extracted information is checked against its source.
- Map interfaces: layout, colours and shapes. Visual assistance does not establish the factual validity of a relationship.
AI-generated content is never treated as factual evidence. Sandbox and digital-twin experiments may use synthetic data; it is clearly labelled and distinguished from observed data.
Our commitments
- Traceable
- Each evidence-based link on a map refers to a verifiable original source. Hypotheses and interpretations are identified as such.
- Transparent
- We tell you where AI was used, and where it was not.
A separate promotional use
The CSR diagnostic available on this website uses AI to generate preliminary content for purely promotional purposes. It is not part of our consulting or accompaniment projects and does not constitute a consulting deliverable. Its generated content must not be confused with the verified evidence used in those projects.
