AP WORKFLOW SOLUTIONS
AP WORKFLOW SOLUTIONS
AP WORKFLOW SOLUTIONS
AP WORKFLOW SOLUTIONS
We’re ten weeks into the year. Some of the industry’s most closely watched forecasts are already playing out — and one is more complicated than expected.
Every January, the Reuters Institute for the Study of Journalism publishes its annual Journalism, Media, and Technology Trends and Predictions report — a survey of 280 news executives across 51 countries that functions as the closest thing the industry has to a shared strategic compass.
This year’s report landed in a particularly charged moment, framing 2026 as the year news organisations would be squeezed from two directions at once: generative AI eating into traffic and distribution on one side, and the creator economy pulling audiences and talent on the other.
Ten weeks in, it’s worth asking: how’s it going?
Some predictions have already moved from forecast to fact. Others are in motion but too early to call. And at least one is turning out to be more complicated than the report suggested. Here’s an honest scorecard.
The report predicted continued, significant decline in search referral traffic driven by AI Overviews and answer engines. This one wasn’t so much a prediction as a confirmation of what was already underway — but the scale has surprised even pessimists.
Google search traffic to publishers declined globally by a third in the year to November 2025, with Google Discover referrals down 21% over the same period. Press Gazette
Publishers now forecast search engine traffic will fall 43% within three years, with a fifth of respondents expecting losses above 75%. Search Engine Land
The pattern is uneven. Publishers specialising in lifestyle or utility content such as weather, TV guides, or horoscopes have seen the steepest declines, linked to the arrival of Google’s AI summaries at the top of search results. Press Gazette Breaking news, counterintuitively, is holding up better — more on that below.
The practical implication the report pointed to — reduce dependency on search, invest in owned audiences, newsletters, and direct relationships, is now less of a strategic choice and more of a survival necessity.
The report predicted AI would shift from a tool newsrooms experiment with to core production infrastructure embedded in workflows. That shift is clearly underway.
The report found that back-end automation was considered important by 97% of respondents, with newsgathering and coding both gaining significant traction. What’s changed since January is less the intention and more the pace — the gap between early adopters and everyone else is widening faster than expected.
The harder question the report raised — around agentic AI, where systems execute multi-step tasks without human initiation at every stage — is still more prediction than reality in most newsrooms. But the infrastructure being built now is what agentic workflows will run on.
The decisions being made today about AI governance are going to matter a great deal when that moment arrives.
"AI is becoming a core infrastructure in the way we produce and distribute content... The industry is moving far beyond ChatGPT as a tool and is building real AI systems." INMA
This one wasn’t in the original predictions and it complicates the narrative of uniform decline. Breaking news is up 103% across all Google surfaces since November 2024, while every other content category — evergreen, landing pages, homepage traffic — is declining. ALM Corp
The reason matters. AI systems are good at synthesising established information but struggle with real-time, rapidly developing events where the facts are still being established. Breaking news, it turns out, is one of the content types hardest for AI to replace — which is precisely where journalistic infrastructure, speed, and editorial judgment still have a clear advantage.
For broadcasters and news organisations built around live coverage and real-time reporting, this is genuinely good news. It validates the report’s broader argument that the organisations which will survive are those doubling down on what makes journalism distinct, not those trying to compete on volume with AI-generated content.
The report found that 76% of news executives planned to encourage journalists to behave more like creators in 2026, with half planning to partner with independent creators for distribution. This is clearly happening — but whether it’s working is too early to say.
What’s visible so far is the intention. Several major publishers have announced creator partnerships and new studio ventures. The Washington Post’s Ripple project, designed to bring independent writers including Substack authors onto its platform, is rolling out further this year. CNN Creators is launching as a new brand aimed at capturing younger audiences with a more informal register.
The harder question is whether institutional news brands can genuinely adopt the tone and format that makes creator content compelling, without losing the editorial rigour that makes institutional journalism credible. Those two things are in tension, and ten weeks in, nobody has fully resolved it.
The report predicted — and in fact helped popularise — the term “liquid content”: stories that are not static but adapt in real time based on platform, context, and audience. The concept is genuinely gaining traction.
The Washington Post’s AI-generated personalised podcast was cited as an early example, reportedly producing more podcast content in three days than in its entire previous history.
But genuine liquid content at scale — where a single story automatically becomes a 60-second video, a push alert, a long-read, a social post, and an audio clip without manual intervention — is still more aspiration than standard practice for most organisations. The workflow infrastructure needed to support it is the gap.
Newsrooms that have invested in story-centric production platforms are best positioned to close it; those still working with siloed, format-first tools are further behind than the trend coverage suggests.
The report flagged AI licensing — deals between news publishers and AI companies to use their content in training data and responses — as a potential revenue opportunity, while noting uncertainty about how significant it would prove. The picture emerging is cautious at best.
Most publishers who have signed deals report that the revenues are modest relative to the traffic losses they are experiencing. The maths does not yet add up. A few large publishers with significant brand value and content libraries are in a stronger negotiating position, but for the majority of news organisations, AI licensing is not the lifeline some hoped it would be.
The more significant development — which the report did anticipate — is the legal and regulatory pressure building around AI companies’ use of publisher content without licensing agreements. That battle is still being fought, and its outcome will shape the economics of the industry for years.
The honest read of the first ten weeks of 2026 is that the structural pressures the Reuters Institute identified are real and moving faster than many expected. The organisations responding most effectively are not the ones waiting for the landscape to stabilise before deciding what to do.
They are the ones that have already made deliberate choices about where their distinctive value lies — and built workflows to deliver it consistently.
Speed still matters. So does depth. Breaking news rewards infrastructure and editorial judgment. Owned audiences compound over time. AI works best when it handles the routine and humans own the story.
None of that is new. But the data from the first ten weeks of 2026 makes it more urgent.
Talk to the AP Workflow Solutions team about AP Storytelling.
