This State of the Industry report, produced in partnership with Piano, explores how publishers are adopting AI on the business side of their organizations, specifically what applications they use, the challenges they encounter and the AI investments they’re prioritizing.
AI tools are rapidly reshaping how media companies operate, from subscription marketing and ad sales to audience development and product management. Publishers are increasingly adopting AI across business functions tied to revenue, audience growth and operations.
A January 2026 report from Digiday+ Research found that publisher adoption of AI tools has more than doubled from 2022 to 2025, reflecting how quickly AI has become embedded in media companies’ workflows and operational strategies as it moves from an experimental role to a more central position in day-to-day workflows.
In this new State of the Industry report, Digiday and Piano surveyed 80 publisher respondents to understand how publishers are adopting AI on the business side of their organizations — not in the newsroom, but across the teams responsible for revenue, audience growth and operations.
Even as AI adoption accelerates across the publishing industry, the findings show an industry still in transition. The majority of respondents to the Digiday and Piano survey (76%) said they are currently piloting or experimenting with AI on the business side of their organizations. At the same time, less than one-quarter of respondents (18%) said their organizations are actively deploying AI tools across multiple business functions, underscoring how early many publishers remain in their AI adoption journeys.
Nevertheless, AI investment is increasing alongside adoption. Seventy-six percent of respondents said their organization’s investment in AI tools for business-side functions has increased moderately over the past 12 months, signaling that publishers are continuing to prioritize AI despite ongoing uncertainty around implementation, governance and ROI.
As publishers increase their AI investments, many are also grappling with how to build the underlying data infrastructure needed to make AI tools effective and actionable, according to Cedric Ferreira, Chief Product Officer at Piano.
“AI doesn’t lack data; it lacks trustworthy context,” Ferreira said. “A strong foundation isn’t a warehouse — it’s the layer that turns events into meaning, connecting what a user did to what content it was, what it cost to reach them and what the business should do next.
“The publishers getting AI right have stopped polishing dashboards and started adding that context — normalized property and metric definitions, rich content properties, market information — and with the right tools, adding this semantic layer is easy to do,” Ferreira continued. “Then an AI model can deliver valuable insights and action.”
01
Publishers use AI to optimize ad sales, subscriptions and data analysis
Currently, publishers’ top applications of AI tools are for ad sales, subscriptions and data analytics. More than three-quarters of survey respondents (84%) said their organization is currently using or actively piloting AI tools for ad sales or ad operations, including yield optimization, proposal generation and campaign management.
More than three-quarters of respondents also said their company is using or piloting AI tools for subscription and membership marketing, including customer acquisition, retention and churn prevention (79%), as well as for data analytics and business intelligence (76%).
The ways publishers are applying AI closely align with the business departments where they are deploying the technology. Ninety-four percent of respondents said their organization is using or piloting generative AI for marketing copy, ad copy or email drafts, while 68% said their company is using or piloting AI for sales enablement, such as proposal generation, RFP responses and prospecting.
Publishers are also increasingly using AI to support data analysis and decision-making. Three-quarters of respondents (75%) said their organization is using or piloting AI capabilities for natural language querying of analytics data, such as asking questions of dashboards, while 63% said their organization is using or piloting AI for automated reporting and insight generation. These findings highlight AI’s growing role in helping publishers automate analytics workflows and support broader business intelligence functions.
As publishers expand their use of AI for analytics and business intelligence, many are also looking to balance operational efficiency with business optimization strategies.
“The choice between automation and optimization is a false one,” Ferreira said. “AI does both at once. Take a paywall template: ‘Build me one’ is automation — a time-saver. Layer in best practices and automated live testing, and the same workflow becomes optimization — the same motion, both outcomes.
“When you’re operating at scale, the complexity just outgrows what a normal team can manage,” continued Ferreira. “And that’s really where AI comes in, not as some magic layer. Dynamic pricing, paywall logic, lifecycle campaigns, you can’t run those at the level of detail that actually moves the needle with people alone. The automation is what makes it viable. That’s the honest framing: it’s not AI or efficiency, it’s that you need AI to get both at the same time.”
02
AI adoption expands beyond chatbots into workflow automation
When asked which types of AI tools their company is currently using, nearly all publisher respondents (94%) said their organization uses general-purpose AI assistants such as Anthropic’s Claude, OpenAI’s ChatGPT and Google’s Gemini.
More than half of respondents said their company uses Claude (65%) and ChatGPT (51%), respectively, while less than half of respondents (42%) said their company uses Gemini.
General-purpose chatbots like these are typically easy to adopt and fast to deploy. While they can be customized and connected to other tools, they are most often used as ready-made interfaces for a wide range of tasks, including content generation, research, summarization and workflow support.
More than three-quarters of publisher respondents (85%) said their organization is using AI tools that automate workflows, such as AI-powered platform Zapier AI and other custom agents. Unlike general-purpose assistants, workflow automation tools are typically custom-built systems powered by large language models and designed to perform specific business or audience-facing tasks using a publisher’s own content, data and rules.
Because these systems often integrate with internal platforms and execute multi-step processes, they generally require more technical investment, customization and ongoing maintenance than off-the-shelf tools.
03
The benefits of AI applications vary widely by business channel
As publishers expand their AI investments, many are beginning to see tangible business results from those efforts. The majority of publisher respondents (89%) said the impact of AI is measurable at their company. When asked what benefits their organization has experienced from AI adoption on the business side, 70% of respondents said improved audience targeting and segmentation was the top benefit.
Other valuable outcomes included improved decision-making through better predictions (66%) and increased revenue from subscriptions, ads or other revenue streams (63%). Those gains closely reflect the areas where publishers have concentrated their AI investments, particularly in ad sales, data analytics and subscription strategy, suggesting AI’s impact has been most measurable where it is most deeply integrated into core revenue functions.
Similarly, when asked whether AI adoption has led to measurable improvements in KPIs, publishers pointed especially to gains on the subscription side of their businesses. Seventy-three percent of respondents said AI adoption has improved subscriber retention and reduced churn, while 70% said AI has improved subscription acquisition costs, and 61% said it has improved subscription conversions.
Within business areas, AI applications had the most impact in the following ways:
Within the subscription and membership marketing segment, AI applications had the most impact on automated lifecycle email campaigns (90% of respondents) and pricing optimization (76%) — underscoring that publishers are prioritizing tools that deliver immediate, incremental gains in conversion and retention.
Automation emerged as the defining use case for AI applications in ad sales and operations, where efficiency gains are immediate and tangible. AI applications had the most impact on automated proposal or media kit generation (79% of respondents); yield optimization and floor price management (57%; and ad inventory optimization, such as balancing ads versus subscriptions (51%).
For example, Hearst is testing how agentic AI can improve processes for its ad sales division, while Thomson Reuters — owner of news agency Reuters — has incorporated agentic AI in its business divisions. And Belgian-headquartered publisher DPG Media has integrated the tools across departments.
Audience development showed less consensus around a standout AI application than subscriptions or ad sales, suggesting that publishers are still experimenting with how AI can best support audience growth. Rather than focusing on a dominant application, publishers are deploying AI across multiple stages of the audience funnel — indicating that audience growth may be inherently more fragmented and require a mix of tools rather than a core AI application.
As publishers incorporate AI applications across business units, compliance and governance are becoming significant considerations. Ferreira cautioned against treating compliance as a secondary concern in AI deployments.
“AI’s power isn’t a license to ignore compliance,” he said. “Publishers that treat regulation as friction to engineer around end up paying twice — once in fines, once when they rebuild the system properly.
“The smarter approach is to design for compliance from the start,” Ferreira added. “Consent and governance belong in the foundation, not bolted on as a separate layer. You can have the full power of AI without the shortcuts. You just have to build the system to take compliance seriously from day one.”
04
Publishers face data quality and traffic referral challenges with AI applications
While publishers are increasingly finding value in AI applications across business functions, they are also running into structural challenges that could shape how quickly those tools scale.
Chief among them are data quality and integration issues. Seventy-five percent of respondents said data quality issues, such as incomplete, inconsistent or outdated data, affect the quality or usefulness of their AI tools, while 64% of respondents said difficulty integrating data sources into AI tools is their greatest challenge.
Data challenges are common for publishers, but they are particularly acute for smaller organizations without a dedicated in-house AI team. Even publishers that outsource AI tools often have difficulty integrating external tools with internal systems due to privacy and compliance restrictions. Half of the respondents said privacy and compliance restrictions limit data availability.
However, not all publishers see data as a major barrier. Nearly half of respondents (48%) said data challenges are not a significant issue for them, suggesting a widening gap between organizations with mature data infrastructure and those still building foundational systems.
For Ferreira, the root issue isn’t how much data publishers have, it’s how much of it they can actually trust. “The dangerous thing about AI is that it always gives you an answer. Feed it partial or inconsistent context, and it won’t tell you it doesn’t know enough. It will just be wrong — confidently. That’s a real business risk,” he explained.
Ferreira’s prescription runs counter to the instinct to instrument everything. “Less is more, at least to start. Rather than trying to measure everything possible, publishers should guarantee that the right things are measured the right way,” Ferreira said. “Then enrich iteratively — layer in business context, content context, market context. And underpin all of it with a semantic layer that AI can reason across. That’s not a nice-to-have. It’s what separates useful AI outputs from plausible-sounding ones.”
But quality context is only half the equation, Ferreira argued. “The other failure mode is having great context and then doing nothing decisive with it. AI that just surfaces recommendations isn’t transformative,” he said. “The goal is delivering the right experience to the right user at the right time — and that requires context, action and AI working as one system, not in silos.”
That means outcome loops that run autonomously: defining a goal, reading the context, taking action, verifying results — at a tempo and scale no human team can sustain alone. “Publishers with mature data infrastructure aren’t just better at analysis,” Ferreira said. “They’re running a fundamentally different operation.”
05
AI-powered search reshapes publisher traffic and audience strategies
Beyond internal data challenges, publishers are grappling with external pressures that are reshaping referral traffic and distribution. Publishers’ biggest concern about AI’s impact on their business in the next two years is AI-powered search, through tools like ChatGPT and Google’s AI Overviews, reducing referral traffic to publisher sites — 85% of respondents selected this.
The shift to AI-powered search has been fraught with challenges for publishers. Analytics and licensing platform Tollbit found that AI chatbots drive significantly less traffic to publishers than traditional Google search — 95.7% less on average as of Q4 2024.
International trade association Digital Content Next, which counts The New York Times, Condé Nast and Vox among its approximately 40 member companies, also found publishers’ organic search referral traffic from Google has declined broadly. The majority of DCN member sites experienced traffic losses from Google search between 1% and 25% as of the summer of 2025.
In addition to referral traffic disruption, publishers are concerned about broader platform dynamics. Sixty-four percent of respondents said they are worried about platform companies like Google, Meta and Apple using AI to further consolidate power, while more than half of respondents (54%) cited the cost of AI tools and infrastructure as their top concern.
As referral traffic from search becomes less reliable, some publishers are shifting their focus toward owned audience relationships rather than platform-driven discovery. “Owned audiences become the entire game,” Ferreira said. “The shift is from traffic capture to relationship orchestration — turning the right signal into the right action at the right moment, on the surfaces you actually control.”
That shift, he added, has three implications. “Owned touchpoints like your website matter more than search and social, because they’re where you can still shape the experience. Revenue has to diversify away from impressions tied to inbound traffic. And segmentation matters more, not less, when total visits shrink,” Ferreira explained.
“The next decade of publisher growth lives in customer management — better serving the engaged and paying readers you already have, using behavioral intelligence for targeting, and AI as the layer that runs it at scale,” he said.
06
AI adoption shifts from experimentation to operationalization
Looking to the future, publishers are prioritizing AI investments that strengthen audience and revenue performance over experimental or enterprise-wide transformations. Publishers said their top priorities for AI adoption in the next 12 months are: improving audience understanding and segmentation (80%); deploying AI for revenue optimization, including dynamic paywalls, pricing and ad yield (70%); and improving data infrastructure to support AI (68%).
The emphasis on data infrastructure alongside segmentation suggests that many organizations recognize that advanced monetization strategies — such as dynamic pricing and yield optimization — depend on better-integrated, higher-quality data. In that sense, the next phase of AI adoption appears to be less about experimentation and more about building the underlying systems needed to operationalize AI at scale across business channels.
Despite their focus on foundational AI capabilities, relatively few publishers are prioritizing scaling AI at their companies. Fewer than 4 in 10 respondents (35%) said that scaling AI and building in-house AI expertise and talent (25%) were top priorities for their companies in the coming year.
This points to an adoption curve that remains largely pragmatic and incremental. AI investments are being driven by specific use cases rather than broad, enterprise-wide transformation, and, at least for the moment, publishers are prioritizing near-term functionality over long-term structural change.
Ferreira believes the survey data reflects something deeper than pragmatism — it reflects the only viable strategy in a market moving this fast. “AI is advancing quickly enough that what takes 12 months to build today could be table stakes in a few weeks,” he said. “That makes long-horizon transformation programs truly risky. You have to ship your way through this, not plan your way through it.”
That means treating AI as a continuous stream of iterative, additive initiatives — and being equally quick to abandon what isn’t working. “The trap is funding a parallel enterprise AI track with its own team and roadmap,” Ferreira said. “That version ends up disconnected from the systems that produce data and revenue, and disconnected AI doesn’t compound.”
The publishers pulling ahead, he argues, aren’t running a bigger AI program. They’re running a better reflex: test, validate, reinvest — in revenue optimization, analytics, prediction — and move on fast when something doesn’t deliver.
As publishers make the transition from AI as an experiment to full adoption across the business, their goal should be to treat it as a connective layer and an acceleration tool built on a strong data foundation, rather than a separate initiative with its own team, roadmap and budget.
It seems likely that platform traffic will continue to decline. In that environment, publishers need to optimize total revenue and convert readers faster and more effectively — delivering the right experience to the right reader at the right time, consistently and at scale. That requires AI and is how these tools become most effective in a media business: not as a separate initiative sitting above the work, but as the operating system running underneath it.
About Piano
Piano is the digital analytics and subscription management platform that empowers businesses to understand their audience, orchestrate journeys and grow revenue. Its market-leading subscription tools enable clients to engage, acquire and retain paying customers, while Piano Analytics delivers clean, compliant data with AI-powered insights for smarter decision-making. The company serves a global client base including the BBC, Deutsche Telekom, Crédit Agricole, Nikkei, The Telegraph and The Wall Street Journal. To learn more, visit piano.io
