How to turn news articles into assets for AI search

AI is prompting the deconstruction of the normal information article. Publishers involved about declining search visibility should adapt their content material distribution workflows to floor in Google’s AI-powered SERP options and LLMs.

AI-powered algorithms are reshaping how information content material will get surfaced, with audiences favoring video in a social-search hybrid expertise. The article isn’t lifeless, however publishers might have to suppose past it and embrace liquid content material.

Nikita Roy captured this shift throughout her presentation at ONA25:

• “The article is not the unit of journalism in an AI-mediated world.”

Roy additionally challenged the viewers:

• “Should you knew nothing about newsrooms, solely that individuals want trusted, verified data, what would you construct with at the moment’s tech?”

What’s liquid content material?

There’s no consensus but on what “liquid content material” means, very similar to the GEO/AEO/AI search engine marketing debate. I just like the definition shared within the Reuters Institute’s 2026 trends and predictions report: 

• “[Liquid content] describes content material or tales that aren’t static however adapt in actual time based mostly on the viewer’s context, location, time, or interplay. AI facilitates this by tailoring content material to particular person preferences. Requires conventional media corporations to maneuver away from authoring ‘articles’ in direction of extra versatile atomic objects.”

Conventional article elements are all nonetheless current and worthwhile:

• Verified data.
• Quotes.
• Information.
• Sources.

As an alternative of locking them in a inflexible article container, the elements are plugged into versatile content material supply pipelines. The worth shifts from the article as a complete to the details and figures inside it.

How multimodal content material suits into liquid content material

Multimodal content and liquid content material are generally used interchangeably. It might be extra helpful to think about multimodal content material as what flows by means of a liquid content material distribution system, powered by two essential elements:

The secret is to align a writer’s topical strengths with the format preferences of its viewers.

A widely known multimodal content material generator is Google’s Gemini Pocket book (previously referred to as NotebookLM). Give it a PDF of a courtroom ruling, a video explainer, or a 2,000-word deep-dive investigative report, and it will possibly convert the knowledge into a wide range of different codecs: briefings, infographics, quizzes, podcasts, and slide decks. 

I’ve had combined experiences changing knowledge evaluation studies into correct infographics, however that’s anticipated with AI expertise. It’s nonetheless a worthwhile software for testing multimodal choices. Right here’s the infographic NotebookLM generated for this text.

Dig deeper: Utility news content: How to win beyond clicks in AI search

Adapting newsroom workflows for liquid content material

To construct liquid content material workflows, a newsroom’s CMS will need to have the potential to transform an article into completely different codecs. This shouldn’t be a totally machine-based course of, nonetheless. Human enter is crucial. 

As an alternative of forcing a narrative to suit a default article format, the story is allowed to dictate the way it’s introduced in a method that can greatest resonate with its supposed viewers. Steven Wilson-Beales recommends asking: 

• “What’s the important seed of the story and what are the most effective codecs that can permit that seed to bloom?”

Reformatting content material will not be a brand new observe, and publishers are already well-versed in headline A/B testing. With AI, they will check codecs in an analogous approach to uncover what resonates greatest with their audiences.

The tougher half is the personalization layer.

Finnish broadcaster Yle has been working on personalization efforts for over a decade. AI expertise can now assist carry a few of its ideas to fruition, similar to serving up story codecs tailor-made to a person’s present want (an audio model for a driver vs. a text-based article for a subway rider).

Publishers are already experimenting with multimodal and liquid content material in numerous methods:

• Sky Information: Revamping workflows to concurrently develop initiatives throughout a number of distribution platforms as an alternative of changing to digital after tv broadcast.
• Die Zeit: Specializing in podcasts as a multiformat alternative.
• AP’s Storytelling software: Adapts tales into a wide range of codecs, from social posts to push alerts.
• The Washington Put up: Launched an AI experiment in late 2025 with a customizable “Your Personal Podcast” wherein the viewers can select its choice of subjects and hosts. Its bumpy rollout shouldn’t be an excuse for publishers to keep away from AI-powered experiments of their newsrooms, as these pioneering initiatives can supply worthwhile takeaways to construct better-quality merchandise sooner or later.

Structuring articles for AI search

Liquid content material is all about flexibility, but it surely nonetheless must be structured in order that Google’s AI search options and LLMs can simply floor and cite it. 

This implies being conscious of how AI bots course of and extract content material, however not writing solely for them. As an alternative, what’s good for bots might be good for a writer’s viewers.

Components to think about:

• Don’t bury the lede. As an alternative, apply the inverted pyramid construction.
• Use NewsArticle structured knowledge.
• Embrace bullet-point summaries on the prime of articles, particularly for lengthy items.
• Add subheaders to arrange and outline sections.
• Establish the important thing questions, classes, or quotes from the article and spotlight them in a quote field or related artifact.
• Leverage inside linking to deepen viewers engagement and broadcast subject authority.

The very last thing information publishers ought to do is pivot from writing for search algorithms to writing for AI algorithms. Good article construction and formatting may help information content material attain each bots and busy human readers. 

Whereas it could appear to be information consideration is shrinking, it could be extra about matching audiences with their most well-liked platform and format. In a world tormented by information fatigue, liquid content material merchandise may very well be a transformative expertise for information shoppers.

Dig deeper: How to optimize news content for today’s social-first Google SERP

New methods to distribute and monetize information content material

Liquid content material gives the media business a possibility to transition from leasing house on third-party platforms to proudly owning content material pipes stuffed with worthwhile, unique knowledge. It’s a pivot that requires clear-eyed objective, focus, and experimentation to change into worthwhile.

I like to make use of a restaurant as a mannequin for what a thriving liquid content material mannequin might appear like in a newsroom. A restaurant gives a menu of ordinary dishes however adapts in line with its patrons’ dietary, price range, and ambiance preferences. 

Equally, a writer’s content material choices (web site, publication, app) function its menu. By adopting a liquid content material workflow, publishers transfer from a hard and fast menu to a dynamic à la carte expertise tailor-made to viewers preferences.

Monetizing writer knowledge

A 2025 FT Methods report proposed “journalism as a service” (JaaS), the place publishers might monetize unique knowledge by way of APIs or licensing. 

Monetary information publishers might have the present edge right here, however well being, science, and sports activities publishers are sitting on a treasure trove of historic knowledge insights. 

Native publications have related alternatives to change into AI-powered group useful resource facilities or, as Splice Media calls the idea, “Nextdoor for machines.”

Alternatives for affiliate content material

Publishers shouldn’t overlook affiliate content material. AI buying options added to the SERPs, paired with updates to Google’s web site popularity abuse coverage, have left some publishers reeling, whereas others search new alternatives. 

Time is working on a data product designed solely for bots, recognizing the rising potential in agentic AI, the place bots make purchases on behalf of customers.

Can publishers regain their affiliate content material footing by changing into a buying bot’s trusted product evaluate supply?

Discovering the fitting distribution platforms

Liquid content material might be directed to stream to platforms with the most effective engagement alternatives. For instance, sports activities viewership is rising on social media, with followers watching spotlight clips and creator content material as an alternative of watching whole video games on broadcast TV. 

Publishers should dig into their analytics instruments to know not solely their social efficiency but additionally how their social posts present up in search. Google is making it simpler to trace by adding a publisher’s social and video platform data to Search Console.

Personalizing content material distribution

Google can be pushing personalization in its AI-powered information search options. Preferred Sources is designed to assist devoted shoppers join with their favourite information publishers and subscriptions. 

Barry Adams says Google’s personalization options quantity to constructing an viewers loyalty ecosystem. Content material that engages a devoted information model follower, similar to in-depth explainers, turns into a lead magnet for a writer’s different worthwhile merchandise, together with newsletters, subscriptions, and apps.

Different promising developments embody Nota, which makes use of AI to assist monetize writer content material as viewers curiosity spikes. Beakon is an AI-powered reader personalization service for monetary information. I count on this nook of the business to develop quickly within the coming years.

Dig deeper: Content alone isn’t enough: Why SEO now requires distribution

The dangers of liquid content material

Most newsrooms (64%) nonetheless develop tales based mostly on the channel vacation spot (web site, print, TV) reasonably than viewers choice (21%), the Future Newsrooms Research 2026 discovered. Publishers might really feel like they’re out of pivots, however danger being left behind by writer manufacturers which can be extra open to experimentation.

When articles are lowered to informational atoms, the lack of that means is an actual danger. We’ve all seen AI Overviews generate misinformation based mostly on slicing and dicing a number of sources right into a Franken-answer. 

I’ve documented Google’s AI rewriting headlines on Discover, offering fictional scores for a sport that hadn’t even began but. Publishers might personal what flows into the pipes, however Google is on the controls of its Freestyle SERP machine, producing an AI slop concoction with a writer model stamped on it.

Specializing in information content material personalization might strengthen the echo chambers that a lot of society has willingly embraced over the past decade. By giving folks wider management over their information consumption expertise, they might select to slender their perspective to at least one that fits their worldview.

Making liquid content material work for publishers

AI search is prompting a change of the normal article format wherein articles change into repositories of versatile knowledge property that energy AI search responses and LLM citations. 

Publishers ought to optimize the precious knowledge they personal to make sure they will monetize their merchandise in a liquid content material workflow. If carried out strategically, this pivot might create much-needed new income alternatives for the media business.

Dig deeper: Hidden gem publishers outperform major media on audience affinity: Study