An ad repository is the public, searchable database of an online platform's advertising that the EU's Digital Services Act requires very large platforms to maintain, covering what ran, who paid, how it was targeted, and how many people saw it.
Under Article 39 of the Digital Services Act, every very large online platform (VLOP) and very large online search engine (VLOSE) must publish an ad repository: a searchable, API-accessible record of every ad shown on the service. Each entry discloses the ad's content, who it promoted, who paid for it, whether and how it was targeted, and the total number of people who saw it, broken down by EU member state. Platforms must keep entries live for the exposure period plus at least one year after. Meta's Ad Library and Google's Ads Transparency Center are the best-known examples, and TikTok runs one too. ChatGPT does not have one yet, even after its August 31, 2026 VLOSE designation.
An ad repository turns a regulatory requirement into a research tool. Before the Digital Services Act, checking a competitor's live ads meant scraping, guessing, or paying for a third-party tool with incomplete coverage. After it, any VLOP or VLOSE has to make that data public by law.
For advertisers, that shifts competitive research from optional to structural. A platform without a repository, like ChatGPT before its VLOSE designation forces one, is a platform you can advertise on but can't research the way you can on Meta or Google. Tools that aggregate ad repositories across platforms exist precisely because checking six separate repositories by hand doesn't scale for a media team running more than one channel.
