What is OAI-AdsBot? OpenAI's ChatGPT ads crawler

The short answer

• If you run ChatGPT ads, allow OAI-AdsBot in robots.txt and in every firewall, CDN and bot-protection layer in front of your landing pages. OpenAI says it is required for ad review and recommends allowing OAI-SearchBot as well.
• If you do not advertise, it should not reach your site, because OpenAI says it only visits pages submitted as ads. Verify the IP address of any request that claims to be OAI-AdsBot.
• It is neither the training crawler nor the search crawler. OpenAI says its data is not used to train foundation models, and ChatGPT search relies on OAI-SearchBot.

What OpenAI says OAI-AdsBot does

OpenAI's crawler documentation and its help article for advertisers describe the agent this way:

• It validates the safety of web pages submitted as ads on ChatGPT, and may visit a landing page after an ad is submitted to check that the page follows OpenAI's policies.
• OpenAI may also use content from the landing page to decide when the ad is most relevant to show.
• It only visits pages submitted as ads.
• The data it collects is not used to train generative AI foundation models.

It identifies itself with this user-agent string:

Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko); compatible; OAI-AdsBot/1.0; +https://openai.com/adsbot

OpenAI publishes its IP ranges at https://openai.com/adsbot.json. The file was available when we checked on 2026-10-21, with a creation time of 2026-05-12, although some articles have said it did not exist.

Does robots.txt apply?

Yes. OpenAI's crawler overview does not mention robots.txt for OAI-AdsBot, which is why some write-ups call the question unclear. Its guidance for advertisers settles it: OpenAI's crawlers respect robots.txt, and if access is disallowed, crawling stops immediately. OpenAI suggests explicit rules like these:

User-agent: OAI-SearchBot
Allow: /

User-agent: OAI-AdsBot
Allow: /

An explicit group matters. A crawler with no group of its own follows the User-agent: * group, so a Disallow there that covers your landing pages, for example /lp/ or /promo/, would also stop OAI-AdsBot. If you keep private paths disallowed for everyone, repeat those rules inside the new groups.

Why landing pages fail review

OpenAI's help article lists the layers to check, usually with your engineering or security team:

1. robots.txt must allow the landing page's path, as above.
2. Firewalls, CDNs and bot mitigation from providers such as Cloudflare or Akamai can block legitimate crawlers, often with a 403 response. OpenAI suggests allowlisting its crawler traffic, ideally by user agent.
3. Human-verification checks such as CAPTCHAs, JavaScript challenges, behaviour analysis and session validation can block the crawler even after the earlier layers let it through. Exempt OpenAI's crawlers where appropriate.
4. Rate limits can trigger when many ads are uploaded at once. Look for 429 responses, and upload ads in smaller batches.
5. The destination should be a directly reachable web page. App-store links, deep links, documents, and pages that need an app, a login or a particular region may not give the crawler enough content to review.
6. Product feeds need OAI-SearchBot to be able to fetch product image URLs. A 403 from the image host can block feed processing.

OpenAI also says OAI-AdsBot is officially verified and allowlisted by Cloudflare, and that you should not count on a manual bypass of crawler validation. Once access is fixed, re-upload or resubmit the affected ads if their status does not update on its own.

Test a landing page yourself

This request shows the status code your server returns to OAI-AdsBot's user agent:

curl -s -o /dev/null -w "%{http_code}\n" -A "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko); compatible; OAI-AdsBot/1.0; +https://openai.com/adsbot" https://example.com/landing-page

A 200 means rules keyed on the user agent let it through. It does not prove the real crawler gets in: a request from your own IP address will not match IP-based rules or bot-verification systems the same way. After submitting an ad, check your CDN or server logs for OAI-AdsBot requests from the adsbot.json ranges and the status codes they received.

If you rely on fixed IP allowlists, OpenAI points to https://openai.com/adsbot.json and https://openai.com/searchbot.json, and warns against relying only on IP addresses you happen to see in your logs, because its infrastructure changes over time.

Ads, landing pages and organic answers

Ads in ChatGPT began with a US test on 2026-02-09 for logged-in adults on the Free and Go plans; Plus, Pro, Business, Enterprise and Education users do not see them. OpenAI announced an expansion to Canada, Australia and New Zealand in March 2026, and on 2026-08-11 said ads had launched in the UK, Mexico, Brazil, Japan and South Korea. OpenAI says ads are labelled as sponsored, kept visually separate from answers, and do not influence ChatGPT's answers.

Two points from OpenAI's advice on creating ads affect your website directly:

• Link to the most relevant page, such as a product or collection page, rather than a generic homepage. The crawler reads that page to judge relevance.
• Keep page metadata accurate. Suggested ad drafts in Ads Manager are prefilled from existing website metadata: an image, a title and a description.

Organic visibility in ChatGPT is a separate system. It depends on OAI-SearchBot, covered in GPTBot vs OAI-SearchBot and ChatGPT-User vs OAI-SearchBot, and you can measure it with the method in how to track brand mentions in ChatGPT. To check a landing page for indexing and rendering problems as well, run the free AI visibility checker.

Sources

• OpenAI: Overview of OpenAI crawlers
• OpenAI Help: Advertiser guidance for allowing OpenAI web crawlers
• OpenAI Help: Create ads for ChatGPT Ads
• OpenAI: Testing ads in ChatGPT
• OpenAI IP list for OAI-AdsBot

Reviewed on 2026-10-21. OpenAI's ads programme is still expanding; check the linked pages for current markets and requirements.