AI will amplify journalism not replace it

What role should AI play in publishing?

A recent webinar hosted by FIPP – the global media network – offered a thoughtful and timely space to explore this often fraught question.

Under the guidance of Alastair Lewis, FIPP continues to demonstrate leadership in convening open, forward-looking industry discussions on how technology is reshaping the media landscape. This session, like many it has run, encouraged a range of perspectives and highlighted just how important it is that publishers are equipped to understand and respond to change.

What stood out in this particular discussion was the level of apprehension expressed by some contributors. There was evident concern about the role of bots – perceived by some as a threat to content integrity, traffic accuracy and business models. Others questioned the ethics of AI systems “reading” and learning from human-written journalism at all.

There was a strong sentiment that editorial content is inherently proprietary and should only be consumed and interpreted by people. A number of speakers promoted the idea that publishers should focus on developing their own internal models, trained solely on their own content, while preventing access to generalist LLMs like OpenAI or Perplexity.

These are real concerns about the new AI giants and their attitude towards publishers’ content – and they’re worth discussing. But it’s equally important that we unpack them with clarity and context.

Take bots, for example. Some major publishers have trialled blocking them entirely, only to experience a dramatic drop in traffic – sometimes up to 50%. This doesn’t mean bots were pretending to be readers. What it actually shows is how vital bots are to content discovery: they’re used by search engines, aggregators and algorithmic recommendation systems to surface and steer people to relevant journalism.

In today’s media ecosystem, bots – and, by extension, AI – are part of the infrastructure that connects content with audiences. Blocking them doesn’t protect journalism; it risks isolating it.

Another theme in the discussion was the idea that journalism is, by nature, handcrafted and proprietary – something that should remain entirely human-driven. But this overlooks how journalism actually functions. Most reporting, especially in specialist and B2B sectors, evolves by monitoring developments, synthesising multiple sources and following up on news spotted elsewhere. Journalists build stories based on context, verification and relevance – not from a vacuum.

In fact, AI has been part of this process for years. Google itself is powered by layers of algorithms and machine learning that help surface stories, track developments and verify information. Many journalists already rely on it as a discovery tool, using keyword-driven search to spot trends, check sourcing or find angles worth pursuing.

AI doesn’t disrupt this model – it reflects and accelerates it. It helps identify themes, surface niche developments and support the fast, responsive work that journalism increasingly demands.

Then there’s the ethical unease around machines “reading” the news. While understandable, it’s important to remember why we publish at all. Journalism has always been about reach – about informing others, spreading knowledge and shaping public discourse. It’s not content written for one person at a time; it’s written to be shared, interpreted and acted upon. Blocking AI from learning from this content cuts against the fundamental purpose of publishing in the first place.

There’s also a key distinction that often gets missed in this debate. AI doesn’t push news – it pulls. It waits for a user to ask. Journalism, by contrast, pushes information – it creates awareness, offers surprise, and sets the agenda. The two are not in conflict; they are complementary. The more intelligently AI can pull from journalism, the more vital human editorial work becomes in setting the original context.

It’s important to stress that the webinar didn’t push any single agenda. It reflected a moment of real introspection in the publishing world and that’s precisely why forums like this matter. FIPP should be credited not just for hosting the conversation, but for helping publishers approach this moment with curiosity, not fear.

It has even embraced AI-powered tools itself, including Noah’s news discovery platform, which uses advanced (but human-led) technology to help surface breaking, relevant stories for media brands. While we’re not an AI company, we’re proud to support FIPP and others in helping journalism evolve with the tools now available.

The broader truth is this: to publish is to share. And to use technology that allows us to share knowledge more intelligently is not a threat to journalism – it’s an evolution of it.

AI isn’t here to replace publishing. It’s here to help it scale its deepest values: insight, relevance and reach.

Ivan Massow is a co-founder of Tomorrow’s Publisher and chief executive of NoahWire