There is an old saying in journalism:
"Never tell people what to think. Tell them enough so they can think for themselves."
For most of modern history that was exactly how the public consumed information. Editors decided what to publish, journalists gathered the facts, opinion writers argued their case, and readers reached their own conclusions.
The internet changed that.
Suddenly anyone could publish. Corporate media no longer controlled the flow of information. Independent journalists, bloggers, citizen reporters and alternative media found audiences that had previously been impossible to reach.
Many believed that represented the democratisation of information.
Today, I wonder whether we are witnessing the beginning of a new era.
Not one where information is controlled by governments.
Not one where newspapers determine the public conversation.
But one where artificial intelligence increasingly becomes the first editor every reader encounters.
A Question I Didn't Expect To Ask
Recently I discovered that a former publishing platform associated with our organisation, Voice Media Global, had been identified within a research paper examining websites that republish content originating from RT (corporate media worldwide share media this is a proactive and established media protocol where sources are cited)
The paper itself was produced by recognised researchers using what appears to be a sophisticated methodology. Readers can examine their work and reach their own conclusions.
That was not what surprised me.
What surprised me was Google's AI Overview.
Rather than simply directing readers to the report, the AI generated a broad assessment describing the website as being associated with a foreign disinformation and propaganda network rather than a legitimate media organisation.
Whether readers ultimately agree with that characterisation is almost secondary.
The moment that stopped me was this:
When did a research paper become an AI-generated reputation?
That question extends far beyond one website.
It reaches into the future of journalism itself.
AI Is Not The Problem
Artificial intelligence is one of the most extraordinary technological developments humanity has ever created.
Properly used, it will transform medicine, education, scientific research, engineering, accessibility and countless other fields.
I like millions and millions of others use AI every day.
It helps me research, write, compare sources, organise information and challenge my own assumptions.
The issue is not artificial intelligence.
The issue is what artificial intelligence is increasingly being asked to do.
There is an enormous difference between helping people discover information and helping them interpret information before they have encountered the original source.
That distinction matters.
The COVID Turning Point
Many people first noticed this shift during the COVID-19 pandemic.
Questions surrounding lockdowns, vaccine mandates, laboratory-origin theories, school closures and treatment options became more than scientific debates.
They became debates about who had the authority to define legitimate information.
Technology companies, governments, researchers, fact-checkers and media organisations all found themselves attempting to navigate an unprecedented information environment.
Supporters argued that stronger moderation was necessary to reduce demonstrably false or dangerous claims during a public health emergency.
Critics argued that some moderation systems blurred the line between suppressing falsehoods and narrowing legitimate public debate.
Regardless of where one stands on those arguments, COVID demonstrated something important.
The battleground was no longer simply the publication of information.
It had become the infrastructure through which information reached the public.
The New Information Chain
Today information rarely travels directly from publisher to reader. Instead it often follows a new pathway. A research paper is published. Journalists write about it. Think tanks cite it. Search engines index it. AI systems summarise it.
Finally, a reader receives not just information, but an interpretation of the information before deciding whether to visit the original source.
The AI summary becomes the introduction. Increasingly, it becomes the first impression. And for many people, it becomes the conclusion.
The New Editors
For decades we debated whether corporate media exercised too much influence over public opinion.
That debate has not disappeared. It has evolved. The most influential editors of the future may not work for newspapers at all. They may be the algorithms deciding which sources are surfaced.
The ranking systems deciding which voices are discovered. The trust frameworks influencing credibility. And the AI systems providing reputational summaries before readers have even opened a webpage.
This is not traditional censorship. In most cases the article still exists. The website is still online.
The publisher can still write. What changes is the context in which readers first encounter that publication.
Context is powerful. Sometimes more powerful than content itself.
The Questions We Should Be Asking
Foreign influence operations exist.
Disinformation exists.
Researchers should investigate them.
Artificial intelligence can help people navigate an overwhelming information environment.
None of those propositions are controversial.
The more difficult questions are these.
Who decides when analysis becomes a reputational label?
How transparent should AI systems be about the sources informing their summaries?
What safeguards exist when complex research is distilled into a few sentences that millions of people may treat as objective fact?
What opportunity does a publisher have to challenge or correct an AI-generated characterisation?
These are not questions about left or right.
They are not questions about RT, CNN, Fox, RNZ, the BBC or independent publishers.
They are questions about the future architecture of public knowledge.
The Next Information War
The twentieth century fought information wars through newspapers, radio and television.
The early internet shifted power towards individual publishers and readers.
Artificial intelligence represents another turning point.
The struggle is no longer simply over what information is published.
It is increasingly about how information is introduced, contextualised and understood before readers encounter it.
That is a profound shift.
Artificial intelligence should empower human judgement, not quietly replace it.
Its greatest contribution should be helping us examine competing ideas more intelligently—not deciding, in advance, which voices deserve our trust.
Because once we allow any technology to become the first editor of public thought, we owe it to ourselves to ask who is, in turn, editing the editor.