ChatGPT ad library vs Google and Meta: the transparency gap

As of July 2026, OpenAI does not publish a ChatGPT ad library, so there is no native equivalent of the Google Ads Transparency Center or the Meta Ad Library. We have catalogued 89,133 placements across 2,655 advertisers in 108 niches, the inventory OpenAI itself does not disclose, into a searchable public library.

What the Google and Meta ad libraries actually do

Google's Ads Transparency Center, launched in 2021, lets anyone search every ad a verified advertiser has run across Search, YouTube, and Display, along with targeting criteria, creative history, and impression ranges. Meta's Ad Library covers Facebook, Instagram, Messenger, and Threads, and it is the more aggressive of the two. Political and issue ads are archived for years per Meta's policy, and any user can search by keyword, advertiser, or topic. Both libraries are required by platform policy, both are public, and both let journalists, researchers, and competitors see exactly who is paying for what and when.

Does OpenAI publish a ChatGPT ad library?

No. OpenAI began testing sponsored answers inside ChatGPT in 2025 and expanded the program through 2026, but the company has not released an ad transparency portal, an advertiser directory, or a public archive of sponsored placements. There is no search box at chatgpt.com that lets you ask, 'show me every ad Mastercard has run this month.' That omission is the gap our library is built to close.

What we track instead

We capture ChatGPT ad placements as they appear in real conversations, then catalog the advertiser, the creative, the triggering query, and the niche. The result is a searchable index that researchers, brands, and media buyers can use the same way they use the Google and Meta tools. Browse the full advertiser directory for the complete roster. Here is the state of the database as of July 2026.

Who is actually buying ChatGPT ads

The leaderboard is dominated by B2B software and infrastructure brands, and the full ranking lives in our advertisers index. Mastercard leads with 3,490 placements, or about 3.9% of everything captured. Monday.com sits second at 2,939, followed by Cloudflare at 2,633. Enterprise platforms like Salesforce, SAP, and ZoomInfo all appear in the top fifteen, which tells you ChatGPT's current ad inventory is heavily skewed toward high-ACV software buyers rather than consumer goods.

Which niches get the most spend

Enterprise and fintech, SaaS and productivity, and real estate dominate the inventory. Enterprise & Fintech alone accounts for 2,802 ads from 952 advertisers, and SaaS & Productivity adds another 2,380. Niche depth matters because it shows how broadly ChatGPT is monetizing across query types: 108 distinct verticals, ranging from roofing and landscaping all the way up to AI Search Optimization. Note that a single ad can be classified into multiple niches, so the slice values below count appearances, not unique ads.

Why this matters for buyers and researchers

Without a public library, ChatGPT advertising is opaque in a way that Google and Meta stopped being a decade ago. Buyers cannot benchmark CPMs against a category average. Journalists cannot answer 'who is sponsoring AI answers about X.' Competing brands cannot see when a rival's creative rotates in. A public index changes the market: it makes pricing observable, creative public, and concentration visible. The same shift happened when Google opened its Transparency Center and Meta expanded the Ad Library, and the same shift is now starting for AI assistants. Our paid intelligence layer adds bidding pressure, niche trends, and historical replay on top of the public index.

If OpenAI will not publish its ad inventory, someone has to. That someone is now a small set of third-party trackers, and ours is the largest.

How we capture this data