Boards exist to exercise independent human judgment. Doesn’t systematising AI into board work erode the very thing — genuine deliberation — that makes a board worth having?

The full executive answer

It is the right worry, and I would put the failure mode even more sharply than you did: the risk is not that AI replaces deliberation, it is that fluent AI output creates the appearance of deliberation — a beautifully structured analysis that the board waves through because it looks finished. Deference to polish is exactly how groupthink already works with well-produced management packs; AI industrialises the polish. So the design question we set ourselves was: does each feature increase or decrease the friction of challenge? That is why disagreement between our two AI engines is surfaced rather than reconciled away — the platform hands the board pre-packaged dissent to interrogate; why every number carries provenance a director can pull in the meeting — challenge becomes cheap; and why the approvals engine forces a named human to own the judgment that analysis becomes decision input — accountability stays personal, in the way the UK Corporate Governance Code’s comply-or-explain philosophy assumes it must.

There is also an honest empirical claim available to your side of the argument: boards have absorbed every prior analytical technology — management accounts, consultant decks, ERM dashboards — and the quality of deliberation survived where chairs enforced challenge and decayed where they did not. The variable was never the tool; it was the chair. Our contribution is that, for the first time, the challenge process leaves a record: which alternatives were examined, where models disagreed, who probed what, who approved. A board that wants to deliberate genuinely gets instrumentation for it; a board that wants to rubber-stamp will rubber-stamp with or without us — but with us, at least the rubber-stamping is visible in the audit trail, which is itself a governance improvement.

And a boundary I hold in the sales process, not just in this room: we position the platform as input to deliberation, never as a substitute for it, and the fiduciary duty of care stays exactly where law puts it — with the humans. If systematised analysis tempts a board toward lazier judgment, the record we keep makes that laziness discoverable. Tools that make poor governance visible tend, over time, to improve it.