Green status is not truth (Decision Truth Audit)

AI decision intelligence depends on data many companies routinely edit out: risks, decisions and dissent.

Early in my pre-sales architecture career, I joined C-level meetings with a familiar briefing: stay positive, explain how good the solution was, and keep risks or anything not green out of the room.

I thought it was part of sales.

Twenty years later, I see the pattern differently. Companies rarely say they value appearance over truth, yet many design governance for exactly that outcome. I have lost count of the risks, issues and non-green lines I was asked to remove because they would not land well with executives. When dashboards reward green and minutes preserve alignment, conformance becomes the operating goal, and the appearance of control replaces the information needed to exercise it.

The carousel is my take on what is needed for AI to genuinely support strategy and portfolio decisions: a Decision Truth Audit, the evidence chain AI needs, three applications—portfolio decisions, risk management and strategic moves—and a narrow pilot.

With an unedited record, AI could help leaders answer three questions:

1. Which green programme is following the narrative pattern of past failures?
2. Which “resolved” risk keeps returning under different language?
3. Which strategic investment is consuming capital while its success criteria continue moving?

That is the missed opportunity. With honest histories, AI could connect communication, decisions and outcomes, surface decision debt, and help leaders intervene before risks become expensive.

But many companies have trained their reporting systems to produce clean minutes, green dashboards and incomplete evidence. AI can analyse that record with extraordinary speed. It cannot recover the truth that executive culture chose not to preserve.

The quality of AI decision intelligence starts with the quality of truth leaders are willing to hear.

If an AI system challenged a green status with evidence, would your executive team investigate, or ask someone to remove the line?

#DecisionIntelligence #AIGovernance #PortfolioManagement