The dangerous comfort of almost knowing

We keep treating missing context as if it were missing data - It usually isn't.

Bad decisions rarely begin with stupidity

Most failures begin with confidence, not stupidity.

Someone sees numbers that look solid. A report that seems clean. A system that appears reliable. A process that has been followed. A dashboard moves from red to amber, amber to green, and the room relaxes.

People do not usually make bad decisions because they are careless. More often than not, they make them because the evidence appears good enough, and then they lean on it.

Then reality quietly diverges

At first the gap is small. A complaint here. An anomaly there. A person saying, “This doesn’t feel right.” But institutions are not naturally good at listening to discomfort, especially when the system is speaking in numbers.

By the time the gap becomes obvious, careers, companies, reputations and sometimes entire institutions have already committed.

The machine looked more credible than the humans

The Post Office Horizon scandal is such a painful example because it was not a story about the absence of information. Subpostmasters across Britain kept saying the accounts did not add up. The system showed apparent shortfalls. Technical defects existed. Human beings were raising the alarm.

But the institutional machine gave more weight to the neat confidence of the system than to the messy reality of the people standing in front of it.

That is the part that should trouble every leader.

Not simply that the technology was flawed; technology is always flawed. Not simply that mistakes were made; mistakes are always made. The real lesson is that the organisation became more confident as the humans became more desperate.

The High Court later found that Horizon had bugs, errors and defects. The data had been there. The warnings had been there. The lived reality had been there. But the connections between them had not been properly made.

What looked like knowledge was only almost knowing.

Almost knowing feels like governance

And almost knowing is dangerous because it feels so professional.

It has documents, processes, audit trails, meetings, and risk owners. It can point to policies and say, “We followed them.”

This is why modern intelligence failure is so hard to spot from the inside. It rarely feels like ignorance; it feels like governance.

Digital identity is the uncomfortable test case

We see echoes of this pattern in the debate around digital identity in the UK.

The resistance is understandable. When governments or large organisations build systems that connect identity to more parts of daily life, the fear of control is not irrational. History gives people every reason to be suspicious of systems created for convenience and later expanded for control.

Privacy is not a fringe concern; it is a design requirement. But the opposite is also true, and we do not say it clearly enough.

Not knowing who is who has a cost

It creates room for fraud, it duplicates records, it slows services and it weakens safeguarding. It allows people, companies and behaviours to hide behind ambiguity, inconsistency and disconnected systems.

The people who benefit most from confusion are rarely the vulnerable. They are usually the organised, the predatory, the fraudulent and the well-advised.

Meanwhile, honest people wait in queues, repeat checks, prove the same thing again and again, or fall through the cracks between agencies that each hold a fragment of the truth.

The real choice is not privacy versus security

So perhaps the real debate is not identity versus privacy, or security versus freedom.

It is bad intelligence versus trusted intelligence.

Bad intelligence overclaims and it hides uncertainty. It turns partial data into authority. It punishes those who question the official picture. It says, “The system says no,” and treats that as an explanation.

Trusted intelligence does something more modest, and far more useful. It connects evidence across silos, and it explains confidence. It reveals relationships and shows what has changed. It makes uncertainty visible. It keeps human judgment in the loop, not as theatre, but as accountability.

The real risk sits between the systems

This distinction matters because organisations today are swimming in information. Dashboards, risk scores, case files, alerts, watchlists, reports, audit trails, CRM records, transaction histories, open-source signals, internal notes - Rarely do they lack data. What they often lack is the courage and discipline to ask the harder question: Do we truly understand how this connects?

Because the most important risks rarely sit neatly inside one system. They sit between systems, between departments, between identities and between behaviours. Between what the data says and what people on the ground already suspect.

Until it doesn’t

The dangerous comfort lies in stopping at “almost knowing.” It feels decisive. It looks professional and it allows everyone to move on.

Until it doesn’t.

And when that moment comes, the question is rarely, “Did we have the data?” - it is usually much harder:

• Why did we not understand what it meant?

• Can we stand behind what we think we know?

  
  

That remains one of the most quietly decisive questions any leader can ask.