3JUN

Publisher AI adoption hits 93% in Q4 2025

3 min read

08:55UTC

Digiday+ Research reported publisher AI adoption at 93% in Q4 2025, up from 42% in 2022. The inflection point is complete; the open question has shifted to public-facing deployment, where roughly half of publishers still keep AI use internal only.

← Back to Lenovo runs football's biggest AI broadcast Jump to analysis ↓

IndustryDeveloping

Key takeaway

Digiday+ Research recorded publisher AI adoption at 93% in Q4 2025 against 42% in 2022.

Digiday+ Research reported on 15 May that publisher AI adoption reached 93% in Q4 2025, up from 42% in 2022, a 51 percentage-point swing across three years that closes the adoption inflection in the publisher segment⁠¹. The figure validates the market context in which BuzzFeed Inc's AI-first restructuring under Byron Allen is now being priced, where AI strategy is the asset rather than the side bet.

Digiday flagged public-facing deployment as the next open question. Roughly half of publishers still keep AI use internal only, which means the next time-series cut is not whether publishers use AI but whether named publishers disclose it to readers and integrate it into the on-page experience. EU AI Act Article 50 will force the same disclosure gradient on European broadcasters from 2 August 2026. The Q4 2025 baseline sets the line against which the 2026 numbers, with Article 50 in force, will be measured.

Deep Analysis

In plain English

Nine in ten publishers now use AI in some form, according to Digiday research covering the period through late 2025; in 2022, that figure was 42%. The change happened quickly because AI tools became cheap and easy to access. Most publishers are still keeping their AI use internal, not showing it to readers or labelling it. European law from August 2026 will require publishers to label AI-generated content, which will force that disclosure question.

Deep Analysis

Root Causes

The jump from 42% to 93% adoption across three years has two distinct drivers that operated at different phases. From 2022 to 2024, the driver was productivity: AI tools for writing assistance, SEO optimisation, content summarisation and metadata generation reduced per-article production costs at the margin without requiring editorial process changes.

Publishers adopted these tools because the cost saving was immediate and the risk was low (internal use only, no reader-facing disclosure required).

From 2024 to 2025, the driver shifted to revenue defence: publishers facing 6-12% annual digital advertising revenue declines (Reach reported 8.1% digital revenue decline in Q1 2026; BuzzFeed reported 19.8% ad revenue decline) adopted AI licensing income as a structural hedge.

The Center for Journalism & Liberty's analysis of the Reach AWS deal and the News Corp Anthropic settlement shows this hedge logic explicitly: AI licensing revenue offsets ad revenue decline rather than replacing editorial investment.

What could happen next?

Meaning
The 93% adoption figure, read against the roughly 50% public-facing deployment rate, implies approximately 180,000 to 200,000 publishers globally are using AI internally but have not disclosed it to readers. EU Article 50 converts that into a compliance obligation for all EU-facing publishers by 2 August 2026. The Allen-BuzzFeed acquisition (ID event-00) and the Bartz settlement (ID event-02) both demonstrate the financial stakes of getting the AI disclosure posture wrong: a $120m acquisition priced on AI strategy and a $1.5bn litigation floor on unlicensed scraping show the upper and lower bounds of the publisher-AI value equation.

Source Landscape

This story draws on neutral-leaning sources

Primary parallel: Smartphone adoption among US adults tracked an almost identical curve: 35% in 2011, 64% in 2015, 85% in 2020, and 97% by 2023 per Pew Research. The inflection from 40% to 90% took five years and was driven by price commoditisation (iPhone 4S at $199 in 2011 vs the original iPhone at $499 in 2007) and network effects (app ecosystems crossing the critical mass threshold).

Publisher AI adoption moved from 42% in 2022 to 93% in 2025 in three years, half the smartphone curve time, because the price entry point (ChatGPT free tier, Copilot bundled in Microsoft 365) removed the adoption barrier earlier in the curve.

Counter-parallel: The newspaper industry's adoption of desktop publishing technology in 1985-1995 reached near-universal penetration while simultaneously commoditising the skills of compositors and typesetters, ultimately enabling a 60% reduction in print production headcount over a decade. The parallel tracks: universal adoption of the tool does not prevent the tool from restructuring the labour market of the people who use it.

The 93% adoption figure creates a market structure question for AI tool vendors: with near-universal adoption, the competitive dynamic shifts from market development (convincing publishers to adopt AI) to market share (winning the publishers already using AI). The build-vs-buy decision now runs at the individual tool level rather than the strategic level.

The public-facing deployment gap, where roughly half of publishers keep AI internal, represents the next revenue opportunity for AI transparency and disclosure tooling. EU Article 50 obligations from 2 August 2026 will mandate disclosure of AI-generated content for all EU publishers, creating a compliance-driven purchase decision for disclosure infrastructure.

Publishers currently without a public-facing AI disclosure workflow will need one within 77 days of today, which is the near-term commercial opportunity in the adoption data.

Consensus view: The Reuters Institute for the Study of Journalism's 2026 Trends and Predictions report, surveying 280-plus digital leaders globally, found that 79% of news publishers are prioritising video and AI tooling, with the gap between AI adoption and AI disclosure as the defining strategic tension of 2026.

The 93% Digiday+ figure aligns with Reuters Institute's directional finding, and both datasets agree on the threshold question: the adoption race is over, and the public-facing deployment race has begun.

Counter-view: Pew Research Center's journalism and media surveys have consistently found a significant gap between executive-reported AI adoption and journalist-reported day-to-day AI use.

In Pew's 2025 US newsroom survey, 71% of editors said their organisations used AI tools, while only 43% of journalists said they regularly used AI in their own work. If the 93% Digiday+ figure reflects publisher-level adoption rather than journalist-level daily use, the effective production-workflow penetration may be significantly lower.

Key tension: Whether the 93% adoption figure represents transformation of production workflows or installation of AI tools that remain marginal to editorial output, a distinction that determines whether the public-facing deployment gap is a policy question (should we disclose?) or a capability question (can we ship it to readers reliably?).

Sources:BusinessWire

Mentions:Microsoft →EU AI Act Article 50 →EU AI Act →Reuters →ChatGPT →BuzzFeed →Article 2 of the US Constitution →Byron Allen →Center for Journalism & Liberty →Reuters Institute for the Study of Journalism →Microsoft 365 →

First Reported In

Update #2 · BuzzFeed sold for AI; agentic ads go primetime

BusinessWire· 17 May 2026

Read original →

Causes and effects

This Event

Publisher AI adoption hits 93% in Q4 2025

Confirms the market context for the BuzzFeed Allen acquisition and reframes the publisher-AI story from adoption to disclosure: the question now is which named publishers ship AI to readers, not which publishers use it internally.

Led to

BuzzFeed bets the company on Branch Office AI apps

Near-universal publisher AI adoption validates the market context in which BuzzFeed's AI-first restructuring and Allen acquisition occurred.

Occurred 19 Mar 2026

Read story →

Penske folds Vox titles into PMX

With 93% of publishers already using AI and search referral traffic draining, Penske acquired Vox's titles as a consolidation wager on owning more titles in a shrinking open web.

Occurred 18 Jun 2026

Read story →

AI news use rises as trust stays low

The Reuters DNR 2026 advances and contextualises the 93% publisher AI-adoption figure with demand-side trust and click-through data, adding a consumer dimension to the supply-side statistic.

Occurred 16 Jun 2026

Read story →

Different Perspectives

Publicis Groupe

Publicis Groupe credited AI demand for its Connected Media growth in the 13 October third-quarter release, reporting net revenue of €3,808m, up 5.0% organically. It named no AI product, so investors get the AI claim without a product to measure it against.

Educational Broadcasting System

Educational Broadcasting System premiered EBS AI Literature Theatre on 16 October, remaking 11 Korean literary classics with generative AI. Korea's public education broadcaster puts AI-made drama on air under its own name, while the BBC's director-general said he 'got in trouble' for praising an AI Doctor Who episode.

Hakuhodo

Hakuhodo began selling its in-house Image Similarity Checker on 19 October, scoring AI-generated ad creative against similar images on the web and logging who checked what. The Japanese agency is selling the compliance step around generative advertising and leaves the legal judgment with its clients.

Lisa Nandy

Nandy told Sky on 15 October she is minded to issue a public interest intervention notice on its purchase of ITV's Media and Entertainment business, on four grounds including news plurality. She notes Comcast's Sky News undertakings expire in December 2028 and do not cover ITV News or ITN, in which Sky would take a 20% stake.

Australian Broadcasting Corporation

The Australian Broadcasting Corporation asked Parliament's AI committee to extend the News Bargaining Incentive to AI firms and hold them to the copyright, defamation and privacy rules news outlets follow. With SBS telling the same committee that robots.txt is routinely ignored, the public broadcasters are contesting the opt-out Anthropic proposed.

Editorial Perfil

Perfil reported on 14 October that Civil Court No. 57 kept its damages suit against OpenAI and Microsoft out of the federal courts, and it wants a pre-trial inspection of their California servers. Its lawyer says Microsoft Argentina came to mediation without a serious offer, and Perfil seeks payment rather than takedowns.

Written with AI from Lowdown's published reporting. Not individually reviewed by an editor. Editorial standards.