Why AI split the room: What media can learn from events
Why the same conference produced two completely different futures
Originally published last week in AI in Media. Reposted here for That Event Stack with light contextual edits.
Recently, at The Definitive AI Forum for Media, Information & Events in London’s Stationers’ Hall, I watched the room divide itself without anyone even having to spell it out. By afternoon tea, it was unmistakable: the media industry on one side, cautious and defensive; the events industry on the other, pragmatic, experimental, almost impatient to get on with the future.
It played out inside a venue that has witnessed almost four centuries of technological panic (from movable type to the first copyright laws), so perhaps it was appropriate that history repeated itself. The morning sessions, dominated by representatives from media companies, carried a familiar unease: the sense that AI was happening to the media, not with it. But when the events executives took over in the afternoon, the tone flipped completely. If the media panels described a disruption, the events panels described a plan.
The morning: an industry aware of the threat but unsure of the path
Publishers weren’t wrong to sound worried. Their content is being scraped, their distribution channels destabilised and their licensing strategies remain undefined. When asked about deals between media organisations and AI platforms, most responses hovered somewhere between “We hope so” and “It’s complicated.” The uncertainty wasn’t feigned; it was structural.
More fundamentally, media leaders struggled to articulate where AI fits inside their business model.
Improve workflows? Yes.
Rewrite journalism? No.
Generate revenue? Unknown.
Protect the archive? Absolutely.
In short, the media industry understands the stakes but hasn’t yet agreed on the strategy. AI feels like a threat they must respond to rather than an engine that could help them grow.
The afternoon: an industry already building with AI
And then came the events world, speaking at the same conference but with the confidence of a sector already halfway up the hill.
Where the morning had been defensive, the afternoon was almost exuberant. Event operators talked about matchmaking engines already up and running; internal LLMs generating hundreds of custom apps; data products with zero customer churn; real-time sentiment scoring; and exhibitor tools that are no longer selling square metres but selling qualified intent.
It wasn’t theory.
It wasn’t futurism.
It was execution.
Informa described a future where exhibitors buy lead intelligence and exhibition access is determined by that relationship. RX explained how they already know what will perform at a New York show based on data from London and Paris. Terrapinn talked about AI not as a threat to face-to-face events, but as the thing that finally makes them scalable.
The contrast couldn’t have been clearer.
In the morning: copyright law.
In the afternoon: commercial architecture.
The real divide: posture
What emerged wasn’t simply that the events industry is “ahead.” It’s that the two sectors have adopted fundamentally different postures.
The media industry has taken a protective stance - understandable, given its assets - but here’s the rub: compulsive protection tends to become paralysis. Meanwhile, the events industry has decided to treat AI as infrastructure, something to build on rather than brace against.
The irony? Both industries share the same pillars: content, audience, data. One sees AI as a threat to those pillars; the other sees it as reinforcement.
And if this divergence continues, the events industry could realistically begin to own the intelligence layer between professional communities and the information they rely on. This is the layer publishers once dominated.
If I were advising a newsroom CEO…
Last week’s version of this piece - written for AI in Media - focused squarely on the media side of the divide. And there, my advice was blunt:
1. Build capability, not caution.
The companies moving fastest don’t have the most policies; they have the most people using AI.
2. Treat AI as a product engine, not a workflow trick.
Ask what you could sell because of AI, not just what you could automate.
3. Get ahead of AI search now.
Discovery is being rebuilt in real time. Visibility will depend on structure, metadata and licensing.
4. Recognise that your next competitor might not be a publisher at all.
Events companies are rapidly building insight and attribution products media companies assumed they would always own.
If I were advising event leaders…
For this repost on That Event Stack, it’s worth flipping the lens.
If the events industry emerged from Stationers’ Hall as the sector most confidently building with AI, then the advice here is less about correction and more about acceleration.
1. Don’t slow down - you’ve already proven the model.
While other industries debate risk, you’re deploying AI at scale. That execution instinct is your competitive superpower (read my Substack about Fishing with dynamite).
2. Treat your data layer as your future P&L.
Prediction, behaviour modelling, exhibitor intelligence: this is where long-term value will sit. Invest accordingly.
3. Industrialise the experiments.
Internal LLMs, dynamic matchmaking, automated content packs — turn what works into infrastructure. Don’t let it stay experimental.
4. Own the ecosystem before someone else does.
You’re positioned to become the central intelligence hub for entire industries. Publishers once held that role. They didn’t keep it.
The split is the story
When I walked out of Stationers’ Hall - the same building where printers once fought over who owned the rights to Shakespeare - it felt like two industries had attended two different events. One was asking how to survive AI. The other was asking how much faster they could deploy it.
The risk for media is not that AI will replace journalism. It won’t.
The risk is that adjacent industries, like events, will use AI to capture the value journalism never had time to reach.
If media doesn’t close that gap soon, the future of insight, influence and audience understanding may be built by someone else entirely.