Perplexity Signs a Multi-Year Licensing Deal With Reuters' Decades-Deep Archive

What the deal actually covers

Reuters and Perplexity announced a multi-year licensing agreement on Monday, and the company that owns Reuters reported it itself: Perplexity can search and display summarized excerpts of Reuters' news archive going back decades, with citations and links pointing back to Reuters.com. Financial terms were not disclosed. [1] [2]

For Perplexity it is the company's biggest publisher licensing deal to date. For Reuters, whose parent Thomson Reuters has done AI licensing before, it is the first such agreement with an AI search company specifically. Reuters CEO Santiago Abascal, who joined in April, called it a step toward sustainable partnerships between news organizations and AI companies, while Perplexity CEO Aravind Srinivas said the deal supports high-quality journalism. [2] [3]

The lawsuit trail behind the handshake

The contrast with Perplexity's recent history is the story. In October 2025, News Corp's Dow Jones and New York Post sued Perplexity over content use, following cease-and-desist notices from Forbes and Wired in 2024 and legal threats from the BBC. The company that publishers took to court for taking their work is now signing the largest archive license in its history, and the shift from defendant to customer happened in under a year. [2] [4]

Perplexity has been building this path deliberately, signing licensing deals with Time, Fortune, The Washington Post, and others while the lawsuits were still pending. The Reuters agreement suggests the strategy of converting legal pressure into paid partnerships is working, at least with publishers who would rather be paid than litigated. [2]

What it means for how AI search cites news

The practical change for users is attribution: summaries of Reuters reporting inside Perplexity will come with citations and links to the original articles, which is the difference between an answer and a replacement for the source. That matters because AI search answers have been criticized for stripping traffic from publishers, and an archive license with mandatory citation is one of the few structures that pays publishers for the past, not just the present. [1] [2]

It also sets a price reference for the next round of negotiations. Every publisher watching this deal now has a data point for what a decades-deep archive is worth to an AI search company, and every AI company without a Reuters-style license has a precedent to answer for. [3]

The privacy angle worth noticing

There is one more layer worth flagging for readers of this site: an AI search firm holding a licensed, decades-deep archive of wire reporting is also holding a structured record of world events that can be queried in ways the original articles never supported. Licensing makes the copying legitimate; it does not change what can be built on top of the corpus. That is not a reason to oppose the deal, but it is a reason to understand it, and it fits a pattern we have covered before: AI companies acquiring or licensing data caches whose value comes from what a model can do with them later. [3] [5]