At the 77th World News Media Congress (WNMC) held in Marseille, France, the most significant realization was not a technological gap but a gap in questions.
As the director of the English news agency AJP under Aju Media Group, I presented examples of artificial intelligence (AI) utilization in South Korea. The country is among the fastest in the world to adopt AI, actively integrating it into newsrooms for translation, automation, and content productivity enhancement. Aju Economy is also preparing agent-based services, liquid content, and predictive user experiences.
However, the concerns faced by global media at the congress were different.
While South Korean media is still asking, "What can we automate with AI?" leading global outlets are questioning, "What should journalism become in the AI era?"
Throughout the three days of presentations and discussions, the focus shifted away from prompt engineering, article summarization, and translation automation. Instead, participants discussed user experience, personalization, agents, information structuring, trust, and the value of original reporting.
AI remained a central topic, but the core of the discussions was not technology; it was journalism.
The most impressive change was the dismantling of article-centric thinking.
The Indian newspaper The Hindu was reconstructing a single article into multiple formats. Readers could view the same article as a 200-character summary, read it in a Q&A format, or listen to an audio explanation. The key was not the article itself but the format preferred by the reader.
Sweden's Bonnier News was transforming its decades-old article archive into an interactive service. Instead of entering keywords into a search box, readers posed questions, and the media's articles provided the answers.
Search was evolving into conversation.
India's Scroll.in took it a step further. They did not view articles merely as reading material but were building a research platform using AI to provide timelines, knowledge graphs, relationship maps, event clusters, and automatically generated Q&A. Their goal was to create a workspace for scholars and researchers to delve deeply into specific issues.
Germany's Ippen Digital showcased an example where over 4,000 articles were automatically produced in a single day during local elections. However, the focus was not on production volume but on personalized news tailored to regions, communities, and ultimately individuals.
Global media were no longer competing in article production.
They were competing in reader experience.
One of the most striking examples at the congress was from the German news agency dpa.
dpa's strategic director Astrid Meyer stated, "News output is not the end; it is the beginning."
The dpa IQ they unveiled represented a completely different concept from traditional news services. While traditional news agencies supply articles, dpa was structuring facts, contexts, relationships, and data within articles in a way that AI could directly utilize.
AI agents could access dpa IQ to request timelines of specific events and retrieve related individuals, past articles, and real-time updates.
It was about facts, not articles.
While many media outlets were contemplating how to produce content using AI, dpa was pondering what journalism should become in the AI era.
A reliable information infrastructure.
Perhaps that is the closest vision for the future of the news industry in the AI era.
Austria's Kleine Zeitung took it even further.
Digital director Sebastian Krause argued that AI should be viewed as a new reader.
For the past 15 years, media have optimized content for Google, fixating on search engine optimization (SEO) and competing for clicks.
But now, ChatGPT, Perplexity, Copilot, and Google AI mode are reading, summarizing, and reconstructing articles.
AI has already become a reader.
As he expressed, there may soon be separate sites for humans and agents.
What they are discussing goes beyond SEO to answer engine optimization (AEO) and generative engine optimization (GEO).
The goal is no longer to rank high in search results.
What information AI cites and which media it trusts is becoming the new competition.
Conversely, as AI advances, the essence of journalism is paradoxically emphasized.
The most applauded speech at the congress was by A.G. Sulzberger, publisher of The New York Times.
Sulzberger did not oppose AI; rather, he advocated for its active use.
However, he reminded attendees that AI companies ultimately grow based on the reporting produced by journalists.
"Most of the facts we know start from someone's original reporting," he said.
Witness testimonies from people met on the ground.
Documents obtained by reporters.
Facts verified through on-the-ground reporting.
AI can summarize that.
It can rearrange it.
But it cannot produce it.
The repeated message throughout the congress was clear.
AI is not a technology to replace journalists.
It is a technology that enables journalists to conduct more reporting.
As AI proliferates, facts become more important.
As synthetic content increases, on-the-ground reporting becomes more crucial.
As information floods in, trust becomes more valuable.
Returning from Marseille, the conclusion I reached was surprisingly simple.
South Korea remains one of the fastest countries to adopt technology in the world. However, rapid adoption does not equate to leadership.
Global media are already discussing the next stage of automation.
From articles to experiences.
From search to conversation.
From content to knowledge.
From readers to agents.
And from automation back to journalism.
As Ezra Eman said, no one has a map right now.
But at least the direction global media is heading is visible.
That direction is not toward AI but toward people.
It is about asking what readers want, how they experience news, and why journalism is still necessary.
Perhaps these are the same questions South Korean media should be asking now—not what to automate with AI, but what more can we report thanks to AI.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.