Publishers Push to Keep Search Access While Setting AI Terms
Watch more: TechReg With Tim Cowen of Preiskel & Co LLP
Tim Cowen is chair of the antitrust practice at Preiskel & Co LLP and co-founder of the Movement for an Open Web.
Highlights
Cowen argues that publishers’ dependence on platforms weakens their ability to refuse content collection.
He proposes separate terms for search indexing and other uses of website material.
He wants regulators to address licensing payments and the distribution advantages platforms give their own AI.
Watch more: TechReg With Tim Cowen of Preiskel & Co LLP
A publisher negotiating artificial intelligence rights with the company that delivers its audience has to consider what refusing the deal could cost.
For Tim Cowen, that dependence explains why Meta’s commercial terms raise a competition question alongside disputes over copyright and contracts. Publishers need access to readers, and a platform controlling that access can dictate conditions extending to material on the publishers’ own websites.
“If you want to get access to that community, then you need to go through that channel,” Cowen, chair of the antitrust practice at Preiskel & Co LLP and co-founder of the Movement for an Open Web, told Competition Policy International (CPI), a PYMNTS company, in an interview.
Cowen characterizes Meta as a monopoly distributor for businesses seeking its social media audience. The concern is that publishers cannot readily substitute another channel reaching the same people. Their dependence gives the distributor bargaining power over content it did not create.
“They’re imposing the distributor’s will on the content creator and publisher, and it’s not really for the distributor to do that,” he stated.
Meta’s proposed terms prompted objections over the collection and use of content from websites associated with its business users. The company removed the reference to training, narrowed the provisions to advertising and business products, postponed implementation and promised consultation with publishers.
Cowen said the Competition and Markets Authority’s inquiry reaches beyond those proposed revisions to existing terms. He regards the investigation as an opportunity to examine what the company already claims the right to do.
“You can’t take the benefit of being a monopolist in these circumstances,” he said. His proposed standard is the bargain that would emerge if publishers had competing distributors and meaningful negotiating power.
Copyright cases address different questions, including whether copying is permitted under exceptions such as U.S. fair use or U.K. fair dealing. Those proceedings can wind up being expensive, prolonged and uncertain, particularly when computing and copying occur across borders.
He cited Getty Images’ dispute with Stability AI to illustrate the jurisdictional difficulties when copying takes place outside the country where a claim is pursued. Publishers must establish what happened, where it happened and which rights apply.
That burden falls unevenly. Smaller businesses confront companies able to finance lengthy litigation while continuing to develop services using disputed material. Cowen also described publishers’ fear of challenging a company that helps determine whether readers find them.
“If you search for something, and it’s not found on Google, then you’ve got a really big problem,” he told CPI. He described the Movement for an Open Web as a consortium through which businesses could challenge conduct they would struggle to contest individually.
Competition authorities offer another route because they can investigate trading conditions and consider interim intervention. Cowen wants those powers used before disputed practices become entrenched.
Cowen’s group has developed website terms permitting search indexing without charge while requiring permission for other uses. Publishers could post prices for material taken without payment and pursue a debt claim.
“So you’re offering a web crawler and a business that’s in the web crawler business the opportunity to take content and pay for it,” he said in explaining the proposal. More directly, he described the intended instruction to a crawler this way. “You can search. If you want to do anything else, ask permission.”
The arrangement uses robots.txt to bring terms to a crawler’s attention, alongside their publication on the website. Meta has pointed to that file as a way for owners to communicate preferences. Cowen’s proposal couples that technical signal with contractual conditions stating what is permitted and what further use would cost.
He said the contract approach was being implemented through publishers and trade associations, with claims moving through the courts. Its practical purpose is to give website owners another means of seeking payment.
Cowen connected that approach to the Like Company proceedings involving Google and a Hungarian publisher. He discussed an advocate general’s opinion addressing the reproduction of publisher material and ways to reserve rights. The opinion is an advisory step in the proceedings.
His underlying argument is that public availability does not surrender the owner’s authority over subsequent uses. That distinction matters as publishers consider how material is collected for training and how it is reproduced or presented to users.
The price, Cowen said, should remain with each website owner. A routine tennis photograph and an image of the winning Wimbledon shot have different commercial values. The latter also has a particular value immediately after the event, when newspapers need it.
“There is a time value to a lot of things in the digital world which really needs to be addressed,” he said.
Distribution charges require a separate assessment. Cowen proposed examining the incremental cost of reaching an additional customer as a benchmark for the charges a distributor could sustain in a competitive market.
He questioned whether an instruction to negotiate adequately addresses the imbalance between a publisher and a dominant platform. Without a workable benchmark, the stronger party retains little incentive to accept different terms.
The Google disputes raise a related bundling problem. Publishers have challenged the use of their content in AI services, and regulators have pursued controls allowing them to refuse AI uses without penalties in ordinary search.
Cowen also objects to Google using search to distribute Gemini while competing AI services must attract users through other channels. “They shouldn’t get the benefit of bundling with a monopoly product,” he said.
He argues that consumers should choose whether to receive AI answers, particularly when they are simply trying to reach a court, government agency or other specific website.
Cowen sees scope for conventional competition law and digital markets legislation, including the U.K. Competition Act, EU Article 102 and the Digital Markets Act. He favors whichever procedure can deliver effective intervention, with fair terms, publisher control and comparable treatment for rival AI services as the objectives.
“[Operating as] a platform that supports other businesses is mutually beneficial,” he said. His proposed open web would preserve that relationship by allowing publishers to be found while retaining authority over further uses of their work.
Watch the full interview with Tim Cowen to learn more:
Tim Cowen is chair of the antitrust practice at Preiskel & Co LLP and co-founder of the Movement for an Open Web.
Get PYMNTS Today, AI, B2B and more.
Own the conversation. Drive demand.
Start a Conversation →