Can a machine do what humans increasingly struggle to do, speak honestly, without fear, favour, or pressure?

The hope is understandable, but it rests on a misunderstanding of what truth really is. Truth has never been just a collection of facts waiting to be discovered. It has always existed in relation to power. Some truths are celebrated, while others are ignored, delayed, or deliberately silenced. In many cases, falsehood does not win because truth is absent, but because truth is inconvenient.

History is full of examples where truth was known yet denied. Wars were justified with fabricated reasons. Environmental damage was dismissed long after scientific evidence was available. Social injustices were normalised through selective data and clever language. Truth did not disappear in these moments; it was pushed aside. Power decided what was allowed to be heard and what was meant to be forgotten.

Artificial Intelligence operates inside this same human world. It does not exist outside politics, economics, or institutions. AI systems learn from books, media, research papers, laws, reports, and online content, all produced by humans. These sources carry biases, silences, and priorities shaped by society. If certain voices are marginalised, AI sees less of them. If certain narratives dominate public discourse, AI learns them as normal. If uncomfortable truths are softened or denied by powerful institutions, AI reflects that pattern.

This is why the idea that AI can independently “take a stand” for truth is misleading. Taking a stand requires resistance. Resistance requires independence. AI has neither. It does not challenge power; it adapts to it. It does not confront authority; it works within boundaries defined by humans (powerfuls). This is not a moral failure of machines. It is simply their nature.

Truth, however, has always demanded something more than accuracy. It has demanded courage. Scientists who warned about climate change faced ridicule and dismissal for decades. Whistleblowers who exposed corruption were punished rather than praised. Journalists who reported inconvenient facts were silenced, jailed, or killed. These people did not merely present data; they stood by it, knowing there would be consequences.

AI never faces consequences. It does not lose a job, go to prison, or risk exile. It does not fear social isolation or economic ruin. Without risk, there can be no courage. And without courage, there can be no genuine stand for truth.

There is another uncomfortable question we often avoid asking, who decides what is false in the first place? In theory, truth should be defined by evidence and reason alone. In reality, it is often filtered through laws, regulations, corporate policies, government narratives, and geopolitical interests. AI systems operate within these filters. When topics become sensitive or politically charged, AI does not break the boundary; it stays inside it. It does not lie deliberately, but it is cautious by design. Truth, on the other hand, has rarely been cautious.

AI is frequently described as neutral and objective. This too is misleading. Neutrality may sound fair, but in a world marked by inequality and injustice, neutrality often favours those already in power. When harm exists, refusing to take sides can quietly protect falsehood. When land is taken unjustly, presenting “both perspectives” without context erases lived reality. When pollution damages communities, balancing science with denial weakens truth. In such cases, neutrality does not serve justice; it blunts it.

Yet rejecting AI altogether would be another mistake. AI has real value. It can make complex evidence understandable, expose internal contradictions in false narratives, reduce emotional manipulation, and support researchers, teachers, and journalists. It can preserve memory against deliberate erasure and help honest voices reach wider audiences. But AI is a tool, not a moral agent. It is a torch, not a compass. It can illuminate the path, but it cannot choose the direction.

The greater danger may not be that AI spreads falsehood, but that humans begin to outsource their moral responsibility to machines. When we say “AI said it,” we avoid asking harder questions. Who trained it? Whose interests are served by this version of truth? Which voices are missing? What truths remain unspeakable?

Truth has never been easy. It has always required effort to seek, humility to accept, and courage to defend. No algorithm can replace that effort. Artificial Intelligence can assist the struggle for truth, but it cannot lead it.

In the end, the real question is not whether AI can stand with truth against falsehood. The real question is whether humans still can, and whether we are willing to pay the price that truth has always demanded.