Imagine you’re responsible for an internationally renowned news app. And with the press of an update button, iOS users worldwide receive misinformation on their home screens — with your trusted brand name attached to it.

This is a real example of the risks news organisations run with AI-based disintermediation; Apple Intelligence’s summary of BBC’s app notifications said high-profile murder suspect Luigi Mangione had shot himself, when in fact he had not.

Similarly, Google has been mocked for responding to basic questions with inaccurate, erratic AI overviews. No matter how robust a newsroom’s internal guardrails may be, AI functionality offered by third parties poses a risk of eroding the trust news organisations have built over decades.

Moving fast without breaking things

We must continue demanding rigorous quality assurance (QA) of new AI functionality by third parties that reversion our content. But when it comes to the experiments we run in our own products, there are many ways to manage generative AI’s risks. Rather than limiting the use of the technology, Schibsted has focused on setting criteria for QA.

As the technology matures, we evaluate where there are significant first-mover advantages to be gained through early experimentation and where it makes sense to simply monitor and evaluate capability improvements.

Whether working with functionality offered by Apple or Google, or with companies that position themselves as journalism-friendly publishing platforms and content marketplaces, all publishers will be faced with major decisions about QA, copyright, and distribution channels.

The choices we make as the information ecosystem evolves require clarity about which decisions are one-way doors with “significant and often irrevocable consequences” and which are easily reversible ways of accelerating learning.

Presently, (legacy) media organisations like Schibsted are in the “AI Efficiency Phase” (as described by Caswell and Fang), where “AI is applied primarily to existing tasks, workflows, and products in ways that essentially operate within the existing competitive environment.”

There are, understandably, fewer efforts in established media companies that act decisively on a future where “the fundamental structure of the news and information ecosystem is different.” While we try to predict and shape what that future might look like, there is tremendous value in learning as much as possible about how to create value today and in the near future.

In December, our leaders across product, design, and tech in Schibsted’s Premium Subscription newspapers met in Stockholm with the Swedish Omni team to dive deeper into AI-scenarios they had mapped out through a Schibsted-wide evaluation of plausible futures.

Rather than addressing the broad societal implications of AI, our group zeroed in on how these changes affect the priorities of our product, design, and tech teams most directly in the coming three or four years.

Here are the most impactful scenarios we discussed and how we can use our strengths to face them.

The atomisation of news

Most Schibsted newsrooms have already baked some AI functionality into tools for news gathering, content curation, and versioning for different audiences and format preferences. We will undoubtedly need to continue rethinking internal workflows to support our ambition of adapting to changing expectations and habits among our users.

We predict generic event reporting will continue to be commoditised, while expert analysis, commentary, and boots-on-the-ground reporting will gain relative value.

Most participants in our discussions agree that even our most celebrated journalistic formats might evolve into something different from what they are today. Although not inevitable, it seems likely that news stories will be further “atomised” into individual components such as facts, quotes, audio, and video clips, ready to be remixed into new versions for particular audiences and format preferences.

This helps enable conversational news stories with a quality and dynamism not previously possible.

Many changes to news storytelling will be enabled by content management systems that bake AI functionality into existing workflows. Schibsted has been fast movers in this field largely because we have built these systems in-house, but we also must be open to the idea that specialised third-party solutions may ultimately challenge our current approach to newsroom tooling.

Revolutionising distribution and versioning

Algorithmic news feeds tailored to users’ needs and preferences are widely implemented through passive and active personalisation systems.

In the future, our personalisation efforts will continue baking editorial signals into content ranking and distribution to effectively combine newsroom judgement with algorithms’ intelligence.

In addition to content distribution, we must create responsible, transparent systems for versioning and formatting of each news story. If we believe traditional articles will eventually lose their function as “the unit of news” and every story can be atomised and versioned in multiple ways, we face fundamental questions about media’s role in recording history.

Without news articles, how will our collective history be told and recorded? How will fact-checking and disputed claims be handled in a world with endless versioning? While we can learn from Wikimedia’s handling of differing opinions and truth, it’s clear content versioning needs stronger guardrails to protect our mission.

For example, verifying an article summary is manageable, but handling multiple versions for different audiences introduces scalability issues with the “human-in-the-loop” principle. While humans make mistakes, AI-generated false information could spread rapidly before proper error-correction systems are in place.

Preparing for a multitude of futures

News organisations must embrace portfolio thinking to prepare for incremental and disruptive changes. While AI may strengthen news destinations, we must also hedge against scenarios where user relationships are disintermediated.

As user behaviours change gradually and then suddenly, we need to prepare for significant tipping points. In that future, users’ trust in our process, legacy, and world-class journalism will likely be monetised differently.

Several scenarios we discussed exist in some form today, yet the speed of change is highly uncertain. While we prepare for changes to the information ecosystem, we boldly put new stuff in front of users to accelerate learning.

With the right guardrails and human-in-the-loop processes in place, we can fundamentally improve the way we produce world-class journalism while ensuring our users can access it in ways that fit their changing habits.

Assuming this is just another version of a story we’ve seen before would be a naïve and risky bet to make.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .