Our Knowledge Team, led by Aakash Chugh, turned its attention to one of the most pressing questions in business today: how AI is rapidly reshaping the corporate world — and how we can use it to our advantage. Silvia Savini and Philip Corsano lead the thought pieces in a five-part series published on illuminem, The Broken Ballot. They unpack one uncomfortable observation: the shareholder vote — invented in the 19th century to approve dividends — is now the mechanism boards rely on to settle questions like multi-decade climate liability, intergenerational risk, and AI oversight, and it simply wasn't built for that job. Every resolution still collapses to For/Against/Abstain, with no way to show how a board weighed the trade-offs behind that choice — which Savini and Corsano argue leaves boards more exposed to fiduciary and litigation risk, not less, however the vote lands. From there, they build an alternative: a six-stage, AI-assisted governance architecture that doesn't replace board judgment but forces it into the open — mapping every stakeholder affected, modelling the scenarios a board is actually choosing between, laying competing ethical frameworks side by side, flagging herd-thinking, and leaving an audit trail of how the decision was actually reached. Part three is the series' most honest turn: it argues that no amount of good process design can manufacture the psychological safety and dissent a board culture needs — “architecture without culture produces compliance without judgment.” The series earns its stakes in part four, stress-testing the framework against a real €4 billion case: INEOS's Project One ethane cracker, approved amid four rounds of litigation from ClientEarth and 14 NGOs, without the kind of stakeholder mapping or scenario modelling Savini proposes. Tellingly, she doesn't claim the framework would have changed the vote — only that it would have made the deliberation defensible. The series closes with a practical rollout: a 2027 pilot on climate-material resolutions, feeding into FCA and SEC stewardship guidance by 2030, and eventually global ICGN standards. This series is particularly interesting for anyone who sits on a board, advises one, or sets policy on ESG, risk, or AI governance; it's the risk map for decisions being made right now. Read the full series → https://illuminem.com/illuminemvoicesprofile/sylvia-savini Article 1 — why the 19th-century shareholder vote can't handle 21st-century governance questions Article 2 — the six-stage AI architecture proposed to fix it Article 3 — why process design alone can't manufacture the culture good governance needs Article 4 — stress-testing the framework against INEOS's real Project One litigation Article 5 — the 2027–2030+ roadmap to turn this into an actual governance standard