The media industry is not short of artificial intelligence experiments. Across newsrooms, AI is already supporting transcription, translation, summarisation, metadata enrichment, content optimisation and other production tasks.
The more difficult question is whether these individual applications are helping news organisations become more distinctive, audience-focused and sustainable — or simply enabling them to perform the same work faster.
The Future Newsrooms Study 2026, produced by FT Strategies in partnership with WAN-IFRA, offers an important benchmark for this discussion. Drawing on responses from 448 newsroom professionals across 86 countries, the study examines how editorial organisations are evolving their strategies, audience relationships, workflows, capabilities and skills.
Its central message is clear: newsrooms broadly recognise where the industry is heading, but many still struggle to translate that understanding into daily operations.
The study identifies four interconnected challenges:
- the Strategy Gap;
- the Audience Trust Gap;
- the Capability Gap;
- and the Skills Gap.
For publishers and news agencies, closing these gaps will require more than adding isolated AI tools. It will require an editorial operating environment in which strategy, content, data, people and technology work together.
1. The strategy gap
Audience engagement is the most frequently selected top priority for newsrooms in 2026, while business sustainability emerges as the most important goal when responses are weighted. However, the connection between strategic objectives and everyday editorial decisions remains inconsistent. According to the study, only 32% of newsrooms demonstrate structured alignment between organisational strategy and editorial coverage. Another 42% show only loose alignment, while 25% remain primarily reactive, with decisions driven mainly by immediate events and editorial instinct.
This is not necessarily a failure of editorial judgement. It is often an operational problem.
A newsroom may have clear strategic goals around audience growth, distinctive journalism or new digital products. Yet those priorities can easily become disconnected from commissioning, resource allocation, content production and publishing when teams work through fragmented tools and separate processes. The practical challenge is therefore to turn strategy into repeatable editorial action.
An integrated editorial ecosystem can support this shift by connecting:
- editorial planning and assignments;
- collaborative production workflows;
- content and digital asset management;
- metadata and classification;
- cross-channel publishing;
- performance and audience inputs;
- and AI-assisted production capabilities.
What do we do about it
For ATC, this is a central principle behind newsasset PLUS. The objective is not to add another disconnected tool to the newsroom stack, but to provide a shared operational environment where editorial priorities can be reflected throughout the content lifecycle — from planning and creation to distribution, reuse and archiving.
2. Audience-first strategies still rely on channel-first workflows
Many media organisations describe their strategy as audience-first. Their production models, however, often remain destination-first. The study finds that 64% of newsrooms still create a story primarily for a specific channel — such as a website, newspaper or television output — before adapting it for other destinations. Only 21% begin with a defined audience need or audience group.
Audience data also frequently enters the process too late. For 45% of respondents, analytics largely function as a lagging indicator reviewed after publication, rather than as an input into commissioning and story development.
This creates a structural contradiction: engagement may be a strategic priority, but the workflow continues to begin with the channel rather than the audience. Moving towards audience-aware production does not mean replacing editorial judgement with dashboards or algorithms. It means giving editors and journalists better ways to combine their judgement with relevant signals before, during and after production.
What can you do about it?
Editorial platforms can support that transition by enabling teams to:
– define audiences, channels and formats earlier in the workflow;
– monitor topics and emerging trends;
– create controlled content variants for different audience needs;
– adapt tone, length and presentation by platform;
– optimise headlines, metadata and social copy;
– and maintain a consistent editorial identity across outputs.
The goal is not indiscriminate content multiplication. It is to make each editorial asset more adaptable while retaining oversight, context and brand consistency.
3. AI maturity must move beyond time saved
The study shows that AI is still being evaluated primarily as an efficiency mechanism. Time saved is the most common measure of AI success, cited by 42% of newsrooms. Efficiency matters. Automating repetitive work can give journalists more time for reporting, analysis, verification and audience engagement. But time saved is not, by itself, a newsroom strategy.
The study proposes a more useful maturity path:
- Efficiency: using AI to perform existing tasks faster;
- Scaling: using AI to produce more of the same output;
- New capability: using AI to enable journalism or audience services that were previously difficult or impossible.
The third level is where AI can create more distinctive value. For example, AI can help newsrooms identify patterns across large information sets, make archives more accessible, support multilingual services, assist verification, personalise content discovery or allow audiences to interact with a publisher’s own trusted knowledge base.
This requires a broader measurement framework. Apart from measuring “how many minutes did the tool save?”, newsrroms should also ask whthere the tools at hand
- enabled more original reporting?
- improved content discoverability and reuse?
- helped them serve an audience that was previously underserved?
- strengthen verification or editorial consistency?
- enhanced engagement with owned content?
- suppported journalists perform work they could not realistically perform before?
How do we deal with this?
Within newsasset PLUS, AI capabilities can support tasks such as summarisation, headline and caption generation, translation, classification, named-entity recognition, metadata enrichment, content recommendations and channel-specific adaptation.
The strategic value, however, comes from integrating these capabilities into editorial workflows rather than treating them as standalone generators.
4. Editorial control becomes more important as AI moves closer to publication
The study reveals a clear pattern in newsroom AI adoption: usage is most established in lower-risk supporting activities, particularly transcription and translation, while more sensitive workflows remain cautious and largely human-led.
Story ideation, drafting, verification and image creation carry direct editorial and reputational consequences. Errors in these areas are visible to audiences and can damage trust in both the journalist and the publication. Responsible newsroom AI should therefore not be designed around uncontrolled autonomy. It should be based on clear use cases, traceable actions and meaningful editorial oversight.
A human-in-the-loop model allows AI to assist with tasks such as:
- proposing headlines or summaries;
- producing initial content variants;
- suggesting tags and entities;
- translating or restructuring material;
- supporting content discovery;
- or highlighting information for further verification.
The journalist or editor remains responsible for reviewing, correcting and approving the output.
This reflects ATC’s broader approach to AI in media: amplifying journalists rather than replacing them.
The role of the technology is to reduce unnecessary production friction and expand editorial capability while preserving the judgement, accountability and standards that trusted journalism requires.
5. Metadata and content infrastructure are foundations of AI readiness
One of the study’s most significant findings concerns the connection between AI maturity and backend infrastructure.
Newsrooms with stronger data and metadata foundations — making content more searchable, retrievable and actionable — report substantially higher levels of AI usage. The study also finds that confidence in technology improves when journalists participate in tooling decisions. This highlights a point that is sometimes missed in discussions about generative AI: an organisation cannot build reliable AI services on top of fragmented, poorly classified or inaccessible content.
Before introducing more advanced AI experiences, media organisations need to consider whether their content is:
- consistently structured;
- accurately tagged;
- enriched with entities and topics;
- stored in searchable repositories;
- connected to relevant rights and usage information;
- available across languages and formats;
- and reusable across brands and channels.
Our point of view
Automated metadata enrichment, IPTC-aligned classification, Named Entity Recognition and Digital Asset Management are therefore not secondary administrative features. They are part of the newsroom’s AI infrastructure.
For publishers and news agencies with extensive archives, this foundation can also unlock new forms of value. Properly structured archives can support internal research, content reuse, thematic products, multilingual publishing and grounded audience-facing AI services.
6. Engagement cannot remain an activity that begins after publication
The study describes a shift from an era dominated by reach and distribution towards one in which community and direct audience relationships become more important differentiators. Yet newsroom time allocation does not fully support that ambition. Reporters are estimated to spend 38% of their working time on production, but only 11% on post-publication activities such as audience feedback, community engagement, performance review and optimisation.
This is a major operational mismatch. Newsrooms say they want stronger relationships with audiences, but journalists often lack the time, tools and integrated processes needed to build them. Automation can help, but only if saved time is deliberately redirected towards higher-value activity. Faster publishing alone will not create community.
A more connected newsroom and audience workflow can support:
- trend and sentiment monitoring;
- performance insights;
- social optimisation and scheduling;
- audience feedback loops;
- multilingual FAQs and subscriber support;
- and conversational access to a publisher’s verified content.
Consider this as an option
Automated metadata enrichment, IPTC-aligned classiA reader-facing AI assistant grounded in the organisation’s own content and archives can help audiences explore a topic, locate relevant reporting or receive contextual answers without relying on an open, unverified information source.
Such services should not be positioned as replacements for journalism. Their role is to make trusted journalism easier to access, navigate and understand.
7. New formats require workflow support, not just higher expectations
Newsrooms are also prioritising a broader range of formats. Short-form video is identified as a focus by 79% of respondents, explainers by 74%, and live events or audience forums by 51%. The opportunity is evident, but so is the production challenge.
Creating separate versions of every story for web, social media, newsletters, video, audio and mobile channels can significantly increase newsroom workload. Without suitable workflows, “multi-format publishing” can become a requirement imposed on journalists without the necessary time, skills or production support.
Technology can reduce this burden by helping teams transform a core editorial asset into controlled, channel-ready components, such as:
- transcripts from audio and video;
- summaries and key points;
- newsletter versions;
- social media copy;
- and platform-specific headlines.
The principle should be create once, adapt intelligently and review editorially.
Cross-channel production should not mean publishing identical content everywhere. Nor should it mean allowing AI to produce unlimited variants without purpose. It should enable each story to be shaped appropriately for the audience, format and channel while remaining connected to a common source of truth.
8. The largest AI barriers are human and organisational
Perhaps the strongest warning in the study is that the main barriers to wider AI adoption are not primarily technical. Respondents identify lack of internal skills or AI expertise (61%), cultural resistance or scepticism (52%;) and unclear use cases or strategic direction (45%).
At the same time, 61% of newsrooms report having no formal training programme for new skills. The study also finds that 57% lack AI representation inside the newsroom. Buying software without addressing these issues is unlikely to generate sustainable transformation.
Successful adoption requires journalists to understand:
- what a system can and cannot do;
- where its outputs require additional scrutiny;
- how it supports a specific editorial task;
- who remains accountable;
- and how its value will be measured.
This means technology providers must do more than deliver product functionality. They must support implementation, onboarding, workflow design and knowledge transfer. Embedded, context-aware support can also help. For example, an internal assistant grounded in platform documentation can allow users to ask practical questions in natural language while working, reducing dependency on generic training sessions and making support more accessible.
However, no assistant can replace a structured adoption programme. Leadership, newsroom participation, clear policies and ongoing skills development remain essential.
Building the future newsroom as an operating model
The Future Newsrooms Study does not prescribe a single model for every media organisation. A global news agency, a regional publisher, a specialist business title and a public-interest newsroom will have different audiences, structures and priorities.
What the findings demonstrate is that future readiness depends on integration.
Strategy must connect with commissioning. Audience insight must inform production rather than merely evaluate it afterwards. AI must support clear editorial objectives. Metadata must make content actionable. Training must accompany new technology. Engagement must extend the workflow beyond publication.
