How AI is Automating Print Production in Publishing

The Shifting Economics of Print Production

Publishers face a persistent economic balancing act. Maintaining physical editions requires navigating costs while print circulation continues to fall. Traditional page layout remains a labor-intensive process. To recapture efficiency, software developers are introducing automation engines capable of suggesting headlines, placing stories, fitting content into templates, and building complete newspaper pages.

During a recent WIZONE webinar, technology providers including Atex, StoryEditor, and EidosMedia demonstrated how automated tools operate. Rather than replacing editorial staff, these systems aim to streamline repetitive mechanical tasks while leaving final gatekeeping responsibilities firmly in human hands.

Balancing Automation with Editorial Control at Atex

Defining the boundaries of software-driven layout is the primary challenge for publishers. James Hirsz, Product Lead at Atex, outlined the operational philosophy during the WIZONE presentation, summarizing it with a straightforward mantra: “AI suggests, editors decide.”

Under the Atex framework, a human editor prioritizes a story, specifies its position and width, and lets the automation engine compute the rest of the page layout. When a headline exceeds its designated space, the system generates alternative options drawn directly from the publication’s proprietary editorial style book. The editor retains full autonomy to accept a suggestion, write a new headline manually, or reject the output entirely.

Hirsz emphasized that print production involves strict spatial constraints that generic AI models struggle to comprehend natively. “A headline has to fit a box without damaging the page. A story has to fit the available area. Images have to work within the layout. Automation therefore has to understand both content and the rules governing its presentation,” Hirsz noted. To maintain structural integrity, Atex requires publishers to codify their style guides into usable rules rather than depending on generic models.

Hirsz outlined three core recommendations for implementing production automation successfully:

• Start narrow, prove the concept, and expand gradually rather than automating everything at once.
• Fix the underlying content model before going live, as automation relies entirely on clean inputs.
• Ground the software in the publication’s specific style guide rather than relying on generic models.

Managing Complex Layouts with StoryEditor

As print design evolves to incorporate pages more complex than a simple headline, image and block of text, publishers require tools that go beyond simple text blocks. Marko Margeta, Director of Product Development at StoryEditor, explained that his company approaches automation through predefined templates and snippets rather than asking AI to invent novel page architectures from scratch.

“If you give everything to AI, I think the result could be sterile or plain,” Margeta warned during the webinar. StoryEditor integrates automation within existing design logic, connecting articles to specific snippets and layout positions. The software automatically selects templates based on metrics such as article size, section, associated images, and information boxes.

When an article exceeds its available space, the system trims unnecessary material to fit the layout constraints. Editors retain the ability to review all automated deletions. “The goal is not to remove every decision from the workflow. It is to reduce the number of decisions that require a person to intervene manually,” Margeta said.

Autonomous Page Building and Visual Identity with EidosMedia

Moving from assistive tools to autonomous generation, Jörg Drees, Director Marketing and Business Development at EidosMedia, demonstrated how AI can assemble complete, print-ready pages from pre-prioritized editorial queues. Once newsroom editors decide which stories run, lead, or are held, EidosMedia’s software handles placement, sizing, headline integration, and image coordination.

“That’s production work, not editorial work,” Drees stated.

To preserve the distinct personality of a publication, EidosMedia’s system analyzes historical production archives and past editorial layout decisions rather than imposing a generic template library. This safeguards a newspaper’s unique visual hierarchy and established relationship between text and imagery. According to data shared during the session, the EidosMedia system operates across more than 60 titles, successfully producing over 46,000 pages in 2026 and saving more than 12,000 hours of layout and quality-assurance labor.

These efficiency gains directly alter how newsrooms deploy human talent. Instead of spending hours manually adjusting text boxes and dragging images, editors can redirect their focus toward quality, investigating whether front-page treatments deserve more attention or verifying that the right image was used. Furthermore, compressed layout cycles allow publications to accept late-breaking news or launch additional sections or special editions without adding the same amount of manual work.

Ultimately, the successful integration of print automation hinges on rigorous preparation. Publishers cannot simply layer artificial intelligence over workflows designed around manual production; the underlying data models, templates, and editorial rules must be systematically defined. While machines execute the mechanics of placement and fitting, the core responsibility of defining news value rests where it has always been—with human editors.