Truth Decay Doesn’t Only Happen on Social Media. It Happens in Data Too.

Chief Justice Sundaresh Menon from Singapore recently warned about the growing problem of “truth decay”, where people form strong conclusions from fragmented, emotional, or incomplete information. His comments were directed largely at social media, where speed, virality, outrage, and selective narratives increasingly overpower context and careful thinking.

Most of us immediately recognise the danger when it appears online. A short clip spreads without context. A headline travels faster than the facts behind it. Partial information quickly becomes accepted as complete truth before anyone slows down enough to ask deeper questions.

But while reading the article, I found myself thinking about something else entirely. Isn’t this also how many organisations deal with data?

Today, many business decisions are increasingly driven by dashboards, KPIs, AI-generated summaries, and rapid reporting cycles. We like to believe that once something becomes “data-driven,” decisions automatically become more objective and rational. But business decisions are still made by humans, and humans are deeply uncomfortable with uncertainty.

That matters more than many organisations realise.

The Human Desire for Certainty

One of the quiet realities of leadership is that leaders are constantly under pressure to reduce uncertainty. They are expected to make decisions quickly, project confidence, reassure stakeholders, explain outcomes, and keep organisations moving forward.

So when a chart appears clear, clean, and convincing, the temptation to conclude becomes very strong.

The dashboard says sales are falling. The KPI shows customer complaints increasing. An AI summary highlights declining engagement.

Naturally, the organisation wants an answer: What went wrong?Who owns the problem?What should we do next?

The challenge is that data rarely arrives with complete certainty attached to it. And yet many organisations behave as though it does.

When Dashboards Start Looking Like Truth

This is one of the quiet dangers of modern analytics.

Professional-looking visuals create psychological confidence. Once information is transformed into polished dashboards, attractive charts, AI-generated summaries, and neatly structured KPIs, people naturally lower their skepticism. The presentation itself begins to feel authoritative.

But presentation quality and conclusion quality are not the same thing.

Behind every dashboard are human decisions about what gets measured, what gets excluded, which timeframe matters, how categories are grouped, and which narrative becomes highlighted. A dashboard may be technically accurate while still producing misleading interpretation.

And this happens more often than many organisations admit.

I once saw a business react aggressively to declining sales trends shown on a dashboard. Marketing campaigns were questioned. Forecast assumptions were challenged. Leadership meetings became increasingly tense.

Weeks later, the root issue turned out to be temporary inventory shortages that disrupted purchasing behaviour.

The charts were correct.

The conclusions were premature.

That distinction matters enormously because once organisations emotionally commit to a narrative, reversing that narrative becomes much harder. Teams begin defending assumptions instead of questioning them.

In many ways, dashboards can accidentally create the same problem we now see on social media: partial information becoming accepted as complete understanding.

AI May Accelerate the Problem

The challenge becomes even more important in the AI era.

Today, generating analysis has become dramatically easier. AI can summarise reports, generate dashboards, identify trends, write narratives, recommend actions, and automate reporting workflows within seconds.

This is genuinely transformative. But easier analysis does not automatically create better judgment.

In fact, the opposite risk may quietly emerge. As information becomes faster, cheaper, and more abundant, organisations may begin moving from “not enough information” to “too much unchallenged interpretation.”

Historically, analysts spent enormous effort collecting, cleaning, organising, and producing information. Scarcity was the challenge. Now, interpretation may become the bigger challenge.

AI can generate more charts, summaries, explanations, correlations, and recommendations than most organisations can realistically evaluate with proper scrutiny. And when outputs appear polished and confident, humans naturally assume the thinking underneath must also be reliable.

But confidence is not the same as correctness.

A beautifully generated dashboard can still lead to poor decisions. An AI-generated insight can still miss operational reality. A statistically correct analysis can still create organisational misunderstanding.

The Quiet Evolution of Analysts

This is why I believe the role of analysts is quietly changing.

Historically, analytics teams were valued largely for producing reports, dashboards, metrics, and visibility. But as AI increasingly accelerates report generation, the value of analysts may shift away from simply producing outputs faster.

Instead, analysts may increasingly become interpreters, challengers, translators, context providers, and decision facilitator. That is a very different role.

Because good analytics is not simply about generating answers. It is about improving the quality of organisational thinking.

Sometimes that means slowing conclusions down. Sometimes it means challenging the obvious interpretation. Sometimes it means admitting uncertainty instead of forcing premature certainty. And sometimes it means helping leaders understand the consequences of acting too quickly on incomplete information.

That may become one of the most valuable skills in the AI era. Not generating more information. But helping organisations think more carefully about the information they already have.

The Real Risk Is Premature Certainty

Perhaps that is the deeper connection between “truth decay” in society and the way many organisations handle data today.

In both situations, the greatest danger is not simply misinformation. It is the speed at which partial information becomes accepted as complete truth.

And in a world increasingly flooded with dashboards, AI summaries, automated insights, charts, recommendations, and endless streams of analysis, the organisations that perform best may not necessarily be the ones that move the fastest.

They may be the ones disciplined enough to pause before turning information into certainty.