How platforms made media dependent, and AI makes strong journalism valuable again
YEREVAN, July 9. /ARKA/. Artificial intelligence is already changing the digital media economy. But contrary to popular thought, it not only exacerbates the crisis of the traditional model built on platform traffic, but also restores value to what has always been the main asset of strong newsrooms: primary journalism, trust, and professional reputation.
For the last twenty years, media have lived by someone else's rules: first social media, then search engines, and now AI platforms. This dependence brought audiences, but simultaneously weakened the most important thing: control over distribution, revenue, and the media's own development logic.
The paradox of this new era is that artificial intelligence, while exacerbating the traffic crisis, is simultaneously restoring value to what the platform era devalued: primary information, verification, specialization, reputation, and quality journalism.
Other people's distribution was taken as their own asset
For a long time, media convinced themselves that the future lay in social media, search engines, and algorithms. Editorial teams integrated this logic into their budgets, processes, and strategies: they expanded their SMM teams, rewrote headlines to fit search results, and redesigned websites as if conforming to someone else's algorithms were the essence of modern media strategy.
The platforms sold a simple idea: strong content would find readers. But this wasn't freedom—it was rent. Their goal was different: to retain users within their ecosystem, collect data, and capitalize on attention. In this framework, media outlets were suppliers of valuable content to a system they didn't control.
The consequences of this model are now clear. According to the Reuters Institute (Journalism, Media, and Technology Trends and Predictions 2026), traffic to news sites from Google Search fell by 33% in a year, and from Google Discover by 21%.
Media managers expect search referrals to decline by another 43% within three years. Twitter/X traffic, according to Nieman Lab citing Chartbeat, has fallen by 70% compared to 2022 levels.
This is no longer just a matter of localized problems for individual editorial teams. The crisis is affecting the very model in which external traffic was considered a universal source of growth.
This is already confirmed not only by research but also by the practices of the largest digital publishers. In July 2026, the British LBG Media (LADbible, SPORTbible, and other brands) warned investors of a sharp deterioration in financial performance following further changes to the Meta algorithm and a reduction in search traffic due to Google's AI Overviews. The company reported a 41% drop in indirect revenue, and its shares instantly lost around 40%. Essentially, the market has seen for the first time such a clear, public example of how dependence on both social platforms and search traffic is transforming from an editorial risk into a direct threat to the sustainability of the media business.
And this is only the first wave of changes. Generative AI can make this dependence even more noticeable.
AI is exacerbating the traffic crisis while simultaneously raising the price of primary journalism
Traditional search left publishers with an opportunity: users asked questions, received links, and potentially navigated to the original source. Generative AI changes the mechanics: it offers not a route to an answer, but a ready-made answer as a product.
According to the Pew Research Center (2025), in Google search results with AI summary, users click on regular links only 8% of the time, compared to 15% without an AI summary. The click rate on links within the AI summary itself is approximately 1%.
Users increasingly receive information without ever reaching the source. Publishers have less and less chance of converting attention into subscriptions, advertising, or other direct revenue.
But AI doesn't threaten everyone equally. It takes traffic away from a weak, dependent, and impersonal media model. And at the same time, it restores value to what the platform era systemically devalued: primary reporting, verification, authorship, specialization, and editorial responsibility. Generative AI doesn't create primary information. It requires source material: facts, texts, data, interviews, and investigations. In other words, it requires the journalistic work of others, which has for too long been perceived as free raw material for the platform-based attention economy.
The market is already beginning to move in this direction. In 2023, ARKA and Novosti-Armenia news agencies signed an agreement with Factiva Limited for additional rights to use their content in Dow Jones services for artificial intelligence-related tasks.
In July 2023, the Associated Press announced an agreement with OpenAI, providing access to a portion of its news archive and technological collaboration.
In 2024, the Financial Times signed a strategic partnership with the same company. Quality journalism is beginning to enter the AI economy not only as a source of traffic but also as a licensed asset.
Why Quality Journalism Is Becoming More Expensive Again
In the new information environment, quality media are becoming significant not only as a platform for reach, but also as a source of context, trust, and a high-quality presence in the digital ecosystem.
While AI systems increasingly rely on authoritative sources, companies increasingly value a presence in strong professional media with a name, standard, and reputational weight.
For brands, this is no longer just a question of PR or visibility. It's a question of the context, the sources, and the quality with which they will be presented in the new digital environment.
What exactly should strong media do?
The mistake of the last twenty years wasn't the use of platforms themselves. The mistake was something else: too many editorial offices began to measure their own value by other people's metrics: not by how much their audience needs them, but by how effectively they serve yet another attention-distribution system.
Today, strong media have a chance to correct this mistake. Not because AI will automatically save journalism, but because the market is becoming more restrictive towards secondary content and dependency.
This requires several specific steps.
First, building a direct connection with the audience—own subscriber databases, email newsletters, podcasts, communities. Anything that isn't algorithmically dependent and can't be curated by the next platform update.
Second, monetizing content as an asset, not just as advertising inventory. Licensing for AI systems, partnerships with data aggregators, archival rights—these are no longer just theoretical concepts, but working models for those willing to use them.
Third, deepening specialization. In a world where machines can quickly recount everything averaged out, the value of what can't be safely gleaned from secondary material rises: expert analysis, investigations, industry analytics, the editorial reputation in a specific niche.
Fourth, stopping treating reach as an end in itself. An editorial team with a hundred thousand loyal readers and direct subscription revenue is more resilient than one with a million random clicks from search engines. AI makes this distinction structural. Fifth, stop viewing platforms as assets. Facebook, Google, TikTok, X, or any other platform are not editorial audiences, but rather a leased channel for accessing them. The LBG Media story has demonstrated once again: an algorithm can change overnight, and with it, a significant portion of traffic and revenue. The only asset that truly belongs to a media company is its brand, audience trust, and direct relationships with readers.
The more intermediaries there are between editorial and reader, the less the editorial team controls its own future. This is precisely why, in the age of AI, direct relationships with audiences are becoming not just an advantage, but a matter of media business sustainability.
AI doesn't eliminate journalism. It eliminates the illusion that it can be dispensed with. In the age of AI, the winners are not those who produce more content, but those who create what AI cannot do without: primary information, trust, and professional editorial responsibility.
Konstantin Petrosov
Director of the ARKA News Agency
