Weekly Digest, Issue #165, 7 March 2025

Vincent Peyregne

07 Mar 2026 — 3 min read

Expanding the Publisher’s Customer Surface in the AI-casting Era.

The shared incentives that once supported quality journalism during the era of navigational and transactional search are now being replaced by AI-native exploratory queries. This change does more than just lower website traffic; it makes the old publisher acquisition funnel outdated.

We are moving away from traditional "Broadcasting" (one-to-many) and "Webcasting" (many-to-many) models, where audiences consume uniform, scheduled artefacts that create shared cultural touchpoints but offer limited personalisation and no agency - the new buzzword. Instead, a disruptive new “many-to-one” paradigm is emerging: what the start-up Alien Intelligence defines as AI-Casting.

Publishers relying on web traffic are currently facing a "monetisation event" failure, as new AI agents that provide direct utility will ultimately threaten to crater legacy revenue. This was starkly illustrated by the "Claude Crash," when the launch of Anthropic’s Claude Coworker triggered a significant market revaluation of information businesses, as investors began to price in the total disruption of the "Search-to-Click" workflow.

Moving from content artefact to architectured knowledge

In this new "many-to-one" era, artificial intelligence acts as the main go-between, extracting expertise out of its original form and delivering it instantly. Instead of speaking to large audiences with fixed content, creators now rely on AI services to turn their work into personalised experiences. These can include essays, podcasts, images, or interactive conversations, all tailored to each person's needs and preferred way of receiving information.

This approach lets a single important insight, such as a journalist’s investigation or a medical researcher’s findings, be reshaped for different audiences simultaneously. Through AI mediation, the same information can become a simple guide for a patient, a technical tool for a medical student, or a summary for a policymaker.

This change is the next step in the digital economy. For publishers, trying to block AI only holds back growth, while finding ways to price AI use is a more practical and forward-thinking strategy.

Resilience and growth now depend on transitioning from defensive "protectionism" to an architectural exposure strategy. Publishers who architect their knowledge to reach customers in new environments will find a significantly larger "customer surface" than those who focus solely on protecting traditional articles. Every piece of organisational knowledge has value, provided it is made available and appropriately priced for the agent era. The path forward is not "data extraction" but "data sovereignty", moving away from static artefacts towards a framework in which expertise is structured to be queryable and priced for agent consumption.

The strategic pivot for the modern publishing house lies in moving from the "content artefact" (the finished book or article) to architectured knowledge. While an article about EU AI regulation is useful to a human in a browser, a structured dataset of those same regulations, mapped by jurisdiction, linked to source documents, and updated in real-time, is a sovereign asset that an AI agent can reason over and integrate into a user’s automated compliance workflow.

By architecting knowledge, publishers radically expand their customer reach. A traditional reader must visit a destination website, but an architectured asset can be embedded directly into a user’s professional workflow via API. This transforms the publisher's value from a destination to a utility, penetrating contexts where a human-centric article would never reach.

What does this look like in practice? Check out Florent Daudens’ three-layer approach to agent-driven monetisation, which features O’Reilly Media's knowledge architecture project created with Miso as a proof of concept to expand publishers’ customer surface with rightful AI.