9. What can AI collect before a heat-pump survey without pretending to design the system or approve a grant?

Draft answer framing: Use AI to prepare information for competent review, not to turn incomplete customer answers into definitive suitability, sizing or eligibility decisions.

Why it matters: MCS’s heat-pump design standard sets requirements for certified contractors undertaking design. Ofgem’s current installer guidance makes application timing and scheme requirements material to eligibility assessment. Neither supports an unrestricted sales chatbot making final technical or funding commitments. (mcscertified.com)

Put this question to the expert:

“Which information is genuinely useful before a survey, which conclusions must wait, and how should the system explain uncertainty to the householder?”

A strong answer should cover:

• Customer-provided property information, existing heating, access constraints, photographs and objectives.
• Labels separating supplied facts, assumptions and information not yet checked.
• Human responsibility for design inputs, calculations, equipment selection and final proposals.
• Country-specific funding routes rather than one purportedly UK-wide grant answer.
• A named owner and review date for policy content; changed rules should trigger revalidation of affected scripts. The government’s standards page provides a concrete example of updates continuing in October 2026. (gov.uk)

Evidence to request: A survey-readiness pack and examples of questions the agent declines to answer definitively.

Expert quote: ____________________