Articles
An erasure request must reach the derived copy An erasure request may affect training extracts, retrieval indexes and stored AI outputs as well as the source record. Map the derived copies, assess their status and record the decision for each relevant location. Shared responsibility starts at the AI service boundary Approve supplier handovers as part of the AI service design. Name who controls data access, supplies incident evidence and restores service, then test that those parties can perform the agreed work. A pricing model needs a learning budget before it needs more features Demand learning can require controlled price variation with an economic cost. Agree the exposure, margin floor and stopping rule before asking a pricing model to optimise revenue. Do not ask a quality model for a score. Ask for an intervention. A quality model can help engineers choose a process intervention. Its commercial value depends on a safe comparison that connects the change to saleable output, stability and cost. Personalise the offer, then prove the margin Personalisation can change which offer reaches an eligible customer. Test that policy with a randomised holdout and measure incremental conversion, margin and later customer outcomes. Automate the close, not the sign-off AI-assisted classification may release accounting capacity, but review and correction work determine the result. Test routine handling against the existing close process while preserving evidence and accountable sign-off. An AI log you cannot retrieve is not an audit trail A logging feature is useful only if the organisation can recover the records it needs. Test access, export, retention and contract-exit retrieval before buying a system that must support consequential decisions. Do not give every demand forecast a human override Human input improves some demand forecasts and weakens others. Define what planners may contribute, preserve the original forecast and test the intervention policy against inventory and service outcomes. Automate the interview script, not the hiring decision A voice agent may standardise early interview evidence, but candidate participation and later work outcomes determine its value. Test information collection separately from scoring, with a person accountable for selection. Give AI the creative queue, not the campaign budget Generative AI can broaden the creative ideas available for advertising tests. Assess approved variants through conversion and later customer quality before deciding which deserve further spend. Faster junior drafting needs an apprenticeship control Better assisted drafts do not establish stronger professional judgement. Assess the capacity benefit while separately testing the unaided review capability the practice needs to maintain. Let AI fill the product-information gap before it rewrites the catalogue AI product descriptions may help where buyers lack decision-relevant information. Prioritise those gaps, generate from verified attributes and test purchase outcomes alongside returns and correction effort.
Bring us the question
Something here already on your desk?
Bring the question, the systems involved and the decision you need to make. We will confirm whether a scoped engagement is appropriate.