I am done apologising for AI's success

For the past three years I have walked into meetings with editorial teams feeling like I needed to justify my existence.

I run a company that provides AI-powered news wire and content services to publishers. We do not replace journalists. Instead we provide infrastructure – scale, speed and coverage density that no newsroom can sustainably produce on current budgets. We help brands fill gaps and build depth.

And yet, more often than not, I have been treated as the threat.

Let me say this clearly. I admire journalists. The ones who report from war zones; the ones who sit with grieving families and find the right words; the ones who confront power when it matters. That work is essential and it is not replaceable. I have never argued otherwise.

What I have argued is that the publications enabling that work need stronger commercial foundations. They need predictable traffic, engaged vertical audiences and scalable content operations. That is where we come in.
I’m sad to say that he resistance has been remarkable.

When publishers trial our AI-generated content for one or two articles a day, those pieces frequently become the highest-performing on their sites. When we are allowed to run an entire section – property, finance, LGBT, specialist verticals – readership typically rises by around 200% within a couple of months. Time spent increases almost in lockstep. In longer engagements, with supportive editors, growth has reached 700%.

At the same time, the same publishers hold emergency meetings about Google cutting referral traffic by 40%. They commission strategy papers and blame the algorithm.

If traffic can rise by 700%, Google is not your primary problem – your internal choices are.

In meetings, there is curiosity and enthusiasm. I am asked intelligent questions. There is a sense that something interesting is happening.

Afterwards, the objections emerge. A tone concern. An SEO hesitation. Legal wants to review. A pilot is narrowed. Another meeting is scheduled. Then silence. The timing is no longer right.

Nobody says no. They simply slow everything down until it dies.

Journalists are trained to interrogate and to find weaknesses in arguments. That instinct is vital when aimed at corruption or corporate spin. Turned inward, against tools that might stabilise their own organisations, it becomes self-sabotage. Stopping me does not stop this shift. It merely delays it until a larger, better-funded player arrives.

Last year I pitched a major London publication. The data was clear. The discussion was constructive.
Everyone in that room has since been made redundant. The publication still exists and is still contracting. It is still looking for solutions. I am still here.

A managing director at a financial title we work with said to me recently, “Less than five per cent of our writers write better than AI.” A senior editor at The Washington Post, who later left in a restructure, put it more bluntly: “Readers prefer the AI content.”

Readers may not say that in surveys. But metrics reveal behaviour, not sentiment. They show what people click, complete and return to. And they are returning for our content.

AI can read every council minute published overnight. It can monitor sector sentiment in real time. It can ingest regulatory filings, earnings transcripts and planning applications, then surface patterns before they become obvious. It can filter global press releases and deliver only those relevant to a given niche. It produces drafts that are structured, cited and scored. That scoring matters.

If you sponsor a national fun run, you can now build a running section – routes, training plans, gear reviews, race calendars. Continuous output at marginal cost, serving readers who care deeply. That audience becomes commercially valuable. Communities form and events, which often out-earn the publication itself, follow. Data products emerge.

Traditional publishing economics required scale. AI changes that. Serving 5,000 highly engaged readers can now make commercial sense because production costs have collapsed.

What AI cannot do is confront a dictator or hold a chief executive to account in person. It cannot exercise moral judgement in complex human situations. No serious person claims otherwise.

The model is simple. AI handles infrastructure – volume, speed, coverage. Journalists focus on work that truly requires human judgement. That is not degradation – it is specialisation.

Interestingly, events companies grasp this immediately. They understand audience economics and do not view technology as an existential threat. They ask when we can begin. As they grow, they hire journalists – for depth, investigation and authority.

For three years I have tried to soften the message. I have acknowledged every concern. I have listened patiently while shrinking newsrooms debated formatting details.

It is no longer ambiguous. Publishers who integrate AI intelligently are growing. Those who delay are shrinking. Larger competitors are entering this space with more capital and less patience.

The jobs I am trying to protect are being lost in the meetings where progress is stalled. That is not coincidence – is consequence.

I used AI to help structure this piece. That is not a disclaimer. It is the point.

Ivan Massow is founder and ceo of Noah