By Marty Swant • December 4, 2024 •
Ivy Liu
New lawsuits and deeper partnerships highlight the delicate balancing act between major publishers and AI companies.
Some of the first media companies to strike deals with OpenAI say they’re getting ready to release new AI-enabled features that aim to benefit readers, publishers and advertisers. One of the latest examples is Dotdash Meredith, which used AI to create a new way to target readers with contextual ads.
Using historical audience data and Amazon shopping data, the company trained a large language model to find correlations between content consumption and potential user behaviors. The new unit is expected to launch today with a major (undisclosed) retailer, Dotdash Meredith chief innovation officer Jonathan Roberts said yesterday while speaking at the AI Trailblazers marketing conference in New York. Other panelists included executives from Time, The New York Stock Exchange and The New York Times.
Trained on 11 billion website visits from last year, Roberts described the new ad unit as “going from a binary Morse code system” with to “a full 4K resolution color screen.”
“You can get an incredibly rich view of what somebody’s going to do,” Roberts said. “Not because of who they were last week on some third party site that you can’t see or interrogate, but because of everyone who’s ever gone through this content before.”
Time is also “just about to launch” its first products developed with OpenAI, said Time COO Mark Howard. Although he didn’t share any details, he mentioned also working with Fox to track content on AI models using blockchain tech and talking with Scalepost about developing marketplaces for micropayments in exchange for the retrieval of publisher content.
Priorities differ between media companies. Joe Benarroch, head of content and media partnerships at the New York Stock Exchange, wants to build a new media network in collaboration with news outlets, martech companies and AI providers. Benarroch, who previously led comms at X, mentioned potential partners like xAI’s Grok platform and martech giants like Sprinklr and Adobe.
Despite the deepening partnerships, some publishers remain concerned about tech companies scraping their content to train AI models and also potentially siphoning readership and ad revenue. Earlier this week, OpenAI COO Sarah Friar told The Financial Times that OpenAI would consider bringing ads to ChatGPT. (The company later issued a follow-up statement saying there were no immediate plans in the works.)
Concerns are also spreading beyond the U.S. into Canada. Last week, a group of major Canadian media companies filed a new lawsuit against OpenAI that alleges ChatGPT’s parent company violated copyright law. Other ongoing lawsuits include complaints filed against OpenAI by The New York Times and The Tribune Company and another against Perplexity filed by News Corp.
Even before LLMs, the Times and other publishers have spent years using machine learning building to build contextual tools for reader personalization and ad-targeting. They hope AI personalization will help them rely less on waning distribution channels like search and social media. That’s why The New York Times is focused on using AI to personalize content across its website, apps and emails.
“In a world in which people are getting the news in the palm of their hand while waiting to cross the street or the traffic light, there’s so many opportunities to reach people and to optimize that audience development,” said Chris Wiggins, chief data scientist at The New York Times.
While all the pressure is on AI companies, publishers also see ways for AI to help pull power away from adtech companies.
“The ad tech companies for a long time said, ‘Oh, don’t you worry about it. We’ve got this, we understand your audience better than you ever will, let us take it from here,’” DotDash’s Roberts said. “That wasn’t true then and it’s definitely not true now.”
https://digiday.com/?p=562243
