Research: Only 1% of newsrooms have fully scaled AI

By Dorinne Hoss

Arc XP

Chicago, Illinois, USA

Connect      

AI is no longer a theoretical disruptor. It’s a business reality.

In a new report The State of AI in the Newsroom: Framing the Impact of AI Beyond Workflow Automation in 2025, Arc XP and Digiday surveyed 108 publishers and broadcasters to understand how media organisations are incorporating AI into their newsrooms.

The report explores how far publishers have progressed with AI initiatives, the challenges they face, and the strategic shifts they’re making beyond mere efficiency gains.

While 97% of publishers are actively investing in AI, very few are truly embracing its capacity to reinvent media.
While 97% of publishers are actively investing in AI, very few are truly embracing its capacity to reinvent media.

The results are striking: 97% of publishers surveyed say they’re actively investing in AI, and three-quarters are directing that investment toward editorial.

But here’s the catch: Investing in AI isn’t the same as evolving with it.

Most media companies are stuck in the first gear, experimenting with workflow automation while individual creators and leading-edge media companies are reimagining what editorial means in an AI-native world.

The real threat to your newsroom isn’t automation. It’s irrelevance.

Content has been commoditised; insight has not

Generative AI can summarise, repackage, and distribute your stories faster than any newsroom can. That means speed, volume, and even exclusivity no longer guarantee differentiation.

What AI can’t replicate yet are your instincts, mission, connection to your audience, and editorial judgment. The Arc XP-Digiday report affirms this reality: Publishers leveraging AI say it’s helping journalists focus on what truly matters — original reporting, local engagement, and audience-driven experimentation.

That’s why 88% of publishers now prioritise using AI to elevate content quality, not just efficiency.

The winning formula? Invest in what machines can’t replicate: voice, context, originality, and trust.

As Joey Marburger, vice president of content intelligence at Arc XP, said, “AI is great at copyediting or writing a headline, but it can’t replace the judgment, tone, and soul of a newsroom.”

Multi-modal is mandatory, not optional

If you’re still publishing 90% text and 10% token audio or video and calling it “multimedia,” audiences see right through that.

AI is widening the gap here, too. Tools that accelerate video editing (82% usage) and image optimisation (87%) are helping publishers meet audiences where they are — on platforms designed for sound, motion, and personality.

Yet nearly half of publishers using these tools report dissatisfaction with the results, citing inconsistency, low quality, or editorial misalignment.

That’s the lesson: Speed without standards is useless. Publishers need systems, not just software, that ensure content looks good, works fast, and stays high-quality.

Rethink the media business — from editorial to engineering

Editorial is no longer a silo. Product, engineering, and newsroom operations must all collaborate around a shared goal: sustainable audience value.

According to the Digiday survey:

  • 93% of AI spending is happening within editorial and content teams.
  • Only 23% report centralised executive oversight of AI initiatives.
  • The most effective implementations involve cross-functional AI teams embedded in the newsroom.

This decentralisation is by design. The teams closest to the content are driving innovation. But if you’re not aligning editorial, product, and engineering around audience and revenue, you’re not building a sustainable media business.

AI is a business weapon, not just an efficiency hack

AI is no longer optional. It’s foundational. The forward-looking media brands we surveyed are using AI to:

  • Automate repetitive tasks (86% report reduced labor time).
  • Accelerate content production (85% report faster creation).
  • Drive personalisation and access (88% use AI to create accessible content; 85% for personalised delivery).
  • Fuel experimentation and editorial depth (76% are exploring new formats and workflows).

Critically, only 1% say they’ve fully scaled AI across their organisation, showing that while nearly everyone is experimenting, true transformation is still rare.

The real risk? Becoming indistinguishable

If every publisher uses the same AI tools, trained on the same data, what happens to voice, tone, and credibility?

This is why 90% of publishers say maintaining a competitive edge is their top concern with AI adoption, followed closely by misinformation (89%) and audience trust (85%).

To stand out, publishers must differentiate how they use AI — not just whether they use it. That means fine-tuning models with newsroom-specific content; embedding human editorial oversight as a principle, not an afterthought; and using AI to enhance, not dilute, journalistic values.

“AI is like a junior reporter,” Marburger said. “If you teach it and guide it, it can support the newsroom. But it can’t replace the newsroom.”

The clock is ticking; evolve — or become obsolete

The AI era is dividing publishers into two groups: those who see it as a threat, and those who treat it as a strategic advantage.

Winning organisations will:

  • Treat AI as a platform, not a plug-in.
  • Align editorial, product, and engineering around shared transformation goals.
  • Prioritise originality, insight, and human voice over scale alone.
  • Build hybrid models leveraging both internal and vendor innovation.

This is hard, but so was the shift from print to digital. And just like then, the first movers will define the future. The question now isn’t whether media companies will adopt AI. It’s whether they’ll use it to stand out or risk becoming obsolete.

About Dorinne Hoss

By continuing to browse or by clicking “ACCEPT,” you agree to the storing of cookies on your device to enhance your site experience. To learn more about how we use cookies, please see our privacy policy.
x

I ACCEPT