Francesco Marconi has spent years at the intersection of artificial intelligence and journalism, shaping how newsrooms use the technology before it gained widespread attention. He now argues that communicators must rethink content structure—not for human readers first, but for the AI systems that increasingly determine what information surfaces online.
The first reader is no longer human
Marconi, a computational journalist who founded AppliedXL, says most news is still produced as if people will be the first to see it. That assumption is outdated. AI agents scan press releases, regulatory filings, and articles thousands of times before any human reads them. These systems don’t process information like people—they extract fragments, ignore narrative, and rank content based on clear, retrievable claims.
Organizations that adapt early gain an advantage. Those that don’t may find their content readable but not usable.
Related: Deep Cleaning Targets Hidden Bathroom Areas
The shift has practical effects. A key detail buried in paragraph six of a press release might never be properly weighted by an automated system. Marconi recommends placing the most important information at the top and using precise language instead of marketing language. The bigger change, however, is more fundamental: content must be broken down into clear, structured, and retrievable pieces. It’s not just about the facts but the context that helps machines understand them.
For journalists, this means rethinking how stories are written. Routine, data-driven reporting—earnings updates, sports results, regulatory filings—is already automated. At The Associated Press, Marconi helped generate thousands of earnings stories monthly from structured data before most newsrooms had an AI strategy. The real challenge in journalism, he says, wasn’t writing. It was knowing what to look for, in which data, and when.
Signal before the story
AppliedXL specializes in detecting early signals in public data—what its founder calls “pre-news.” These are updates in clinical trial registries, regulatory filings, or carbon credit issuances that haven’t yet been linked to a story. The time between when a signal appears and when it becomes a headline can range from hours to weeks.
At The Wall Street Journal, his team built tools to monitor large datasets—regulatory filings, transcripts, press releases—to identify signals and break stories ahead of competitors. The goal wasn’t to replace journalists but to assist them, handling routine tasks so reporters could focus on deeper work. Today, AppliedXL turns the expertise of journalists and specialists into AI systems that perform the same steps across millions of data points in seconds.
Related: SoCal Housing Market Sees New Momentum
For communicators, the implications are straightforward. The era of optimizing content only for human readers is ending. AI systems now act as gatekeepers, processing information in their own way. A well-written press release might sound good to a person, but if the key details aren’t structured for machine consumption, they may never reach the intended audience.
Marconi’s advice is direct: “Structure is now more important than prose.” Those who adapt first will have an edge over others still writing for an audience that may no longer come first.
The need for verified information is rising, but the supply isn’t keeping pace. AI-generated content now makes up over half of new web articles, while U.S. newsrooms have lost about half their journalists since 2008. The investigative teams, regulatory reporters, and fact-checkers—those responsible for confirming truth—have been reduced or eliminated.
Related: Media Relations Shifts Focus to Influence
The gap creates problems. Markets trade on unverified information. Courts accept it as evidence. Medical decisions rely on it. As platforms like Google and OpenAI decide what counts as credible, truth is increasingly defined by engineering choices. The tools to verify information exist, but no one has found a way to deliver them at the speed required.
For journalists and PR professionals, the most valuable skills now involve understanding data and specific fields. General knowledge is becoming less valuable, but specialists in pharmaceutical regulation, carbon markets, or regulatory filings remain essential. The challenge is recognizing the type of journalism at play: utility journalism, where structure matters most, or community journalism, where storytelling is key.
Those who fail to see the difference will face the biggest struggles ahead.
