WHEN print operations decouple from digital newsrooms, they face an immediate content gap.
Digital teams produce dozens of stories daily, but not all of them are suitable for print.
This may be because digital formats don’t translate well to print, the topics don’t suit the print audience, or, conversely, content relevant to print readers is not produced at all because it performs poorly on digital platforms.
Moreover, the total volume of stories produced in the newsroom based on the requirements and strategy for digital platforms is often insufficient to “fill the paper” properly.
Meanwhile, print subscribers expect comprehensive hyper-local coverage, community news and the kind of contextual reporting that justifies premium subscription prices.
However, cost pressures have forced many publishers to reduce their local footprint significantly.
There are often simply no boots on the ground anymore to find and provide this kind of content.
A small print team, consisting mainly of page producers and layout editors, cannot generate enough original content using traditional methods and rebuilding local reporting staff is, in most cases, no longer financially viable.
- Teachers, other civil servants face off
- Public relations: Which innovations are driving PR in 2022 and beyond?
- Veld fire management strategies for 2022
- Vitalis Takawira Jr tracing dad’s footsteps
Keep Reading
The solution lies in AI systems, often called AI agents, that can actively monitor, gather, filter, evaluate and adapt content streams, enabling a small team to effectively serve print readers’ unique needs and preferences.
These systems are not only for supporting print production, nor only for regional media houses.
AI agents and AI assistants can, of course, also be used by the digital newsroom as important time savers, and speed and quality boosters, albeit with perhaps different configurations.
The principles and benefits are the same.
I would like to discuss four of them that European publishers such as Schibsted, JP/Politiken, DPG Media, and others are exploring — or actively using — to support both the editorial process in general and print production in particular.
News monitoring and discovery
One very time-consuming part of the day in any newsroom is the “reading in” phase in the morning.
This is when staff spends time checking competitor Web sites (and e-papers), wire agencies, social media and government or municipal sites to see what is new and where interesting stories can be found.
AI agents can do this job much more effectively.
They continuously monitor dozens of information sources simultaneously, far beyond what human staff could track, such as:
Local government feeds: City council agendas, meeting minutes, planning commission filings, budget documents, public notices, etc.
Emergency services: Police reports, fire department logs, court filings, traffic incidents, etc.
Private clubs: Announcements or reports of special interest clubs, sports clubs, community clubs, etc.
Community organisations: School board announcements, non-profit news, church bulletins, etc.
Business intelligence: Building permits, business license applications, real estate transactions, business openings and closures, etc.
Social media monitoring: Local Facebook groups, Reddit posts, community forum discussions, etc.
More advanced AI agents can identify emerging stories by detecting patterns across multiple sources.
They can also detect unusual spikes in local social media activity around specific topics, cluster similar topics or discussions across different platforms or identify correlations between government announcements and community discussions, which suggests when routine events might become newsworthy.
Additionally, they can evaluate the content using scoring algorithms and conduct a news value assessment that considers, for instance, local impact, exclusivity, reader interest or deeper reporting potential.
While setting up these monitoring agents is not always without challenges, it can be a significant time saver, even if only parts of the manual work can be automated.
Incoming e-mail management
While the news monitoring and discovery agents are actively searching for potentially interesting information, newsrooms also receive hundreds of e-mails daily: Press releases, tips, event announcements, letters to the editor, complaints and comments and other general correspondence.
This is equally time-consuming. In some newsrooms, skilled journalists do this task because, again for cost-saving reasons, assistants who used to do this in newsrooms were eliminated.
AI can help automatically filter, categorise, and prioritise these communications. It can also forward them to the right person or team in the organisation.
These include:
Alerts about breaking news and time-sensitive tips.
Community events like meetings, celebrations, village fêtes, and concert or festival announcements.
Business news about new openings, closures, economic developments, pub events and company press releases.
Government communications such as council announcements, policy changes, and public meetings.
Routine updates like standard press releases, promotional content, and non-urgent information.
These systems could also automatically acknowledge receipt of tips, request additional information using intelligent questionnaires, and schedule follow-up reminders for time-sensitive stories.
This helps maintain community trust and increase the likelihood of follow-up engagement.
As with the news monitoring and discovery systems, the implementation has its challenges when it comes to analysing e-mail content, extracting the text and non-text information accurately, filtering the right e-mails, and allowing the system to make the most accurate assessments possible.
However, as AI systems get better and experience in creating AI agents grows, it can make a big difference in a newsroom.
AI-powered content drafting
With the results of the discovery or e-mail filtering agents, GenAI agents can generate first drafts human editors can review and enhance, such as council meeting previews with agenda points, event announcements with key facts, business updates such as restaurant openings, community updates or police incident reports.
These are usually very short pieces and can be used as news briefs in the print edition in side columns.
They are also relevant for many print readers so they feel informed about what is going on, while requiring minimal human intervention to produce.
Template-based article automation
Another option for streamlining the production of content in general, or for the print edition specifically, is template-based natural language generation (NLG).
This is a safe way to automate content creation as it uses structured data and prewritten templates with rules attached to create articles.
This has been around for a while now, but is still not widely adopted in newsrooms. Bergens Tidende for AP and The Wall Street Journal, for instance, use it to create routine articles such as sports recaps, real estate reports, weather or financial reports, where structured data is available.
With GenAI and automation gaining more attention in newsrooms, template-based NLG and machine learning-based NLG, which utilise large language models and machine learning (or a hybrid of both) to create articles, will increasingly become part of digital and print news production.
They can also help teams create content pieces with little human involvement.
The path forward
AI-powered content generation and adaptation can solve the fundamental challenge facing decoupled print operations that don’t have their own content creation staff: creating enough high-quality, locally relevant content with minimal staff.
By automating routine discovery, evaluation and drafting, these systems free journalists to do what they do best: apply editorial judgement, contextual knowledge and refine drafts to suit the interests of their readers.
The result is a print operation that does not just survive separation from digital; it can rebound, offering print subscribers more comprehensive coverage and deeper community engagement once again.
The technology exists today, albeit still maturing.
The real question is which print operations will be bold enough to embrace it, start experimenting and build organisational confidence in using AI agents.
Start with one agent and one use case.
Build trust in its output and scale from there.
AI agents won’t replace editors.
Rather, it can give them back the time and space to do what they do best.
Dr Dietmar Schantin is the principal at IFMS Media Ltd. in London, United Kingdom, and Graz, Austria. He is also co-founder of the AI-collective. He can be reached at d.schantin@ifms-ltd.com or @ifmsMedia.
