And the final speaker in this session at the 2026 Association of Internet Researchers conference in Mexico City is the fabulous Raul Ferrer-Conill, whose focus is on the use of news in AI formats. This represents a new form of platform pressure, similar perhaps to earlier pressures to have news presented on social media platforms: now AI products are also being used to access news content, and news providers are forced to respond to this.
To date, news has appeared in AI largely because Ai companies have scraped news content from news Websites; this is governed (at least in principle) by the robots.txt file on Web servers that specifies what kind of data access sites allow. News companies have increasingly blocked content scraping for AI purposes (though AI crawlers have often ignored such restrictions), and increasingly news companies have begun to licence their content to AI companies instead. The specific licencing deals have been highly secretive, and some such agreements have been breached and become subjects of lawsuits between news publishers and AI providers by now.
The presence of news in AI content has already substantially reduced click-through rates to the actual news articles; this produces significant business problems for news publishers that rely on audience numbers. What, then, is to be done here?
The focus here is on the Internet Engineering Task Force, which maintains and develops the technical standards of Internet infrastructures. The IETF now has a working group on AI preferences and protocols that are intended to describe content access and usage rules, and its work is being examined here through an analysis of its working documents and the stakeholders involved in them.
One of the key areas of contestation is between the differences between search and summary: content can be used to train AI models, used by AI to summarise an issue, or presented directly to the user in response to a search query. These discussions are dominated by technology companies, however, with news organisations and other bodies much less represented. News organisations are outpowered and outnumbered here.
The outcome of these drawn-out negotiations between search and summary definitions will shape journalism’s future economic model, however, and determine whether journalism remains the primary gatekeeper of publicly relevant information; standards organisations like IETF rather than legislative and regulatory bodies have an outsized impact on these decisions, which is problematic from a civil society point of view.
