Which email verifiers block AI crawlers?
Few do, and they block different bots. Of 37 email verification vendors checked on 24 September 2026, only NeverBounce blocks ChatGPT's search crawlers in robots.txt. emailverify.io, Bounceless and Xverify block AI training crawlers at the firewall while robots.txt allows them. No vendor blocked Claude-User, Claude-SearchBot or Perplexity-User in robots.txt.
The short version
- One vendor of 37, NeverBounce, disallows OAI-SearchBot and ChatGPT-User in robots.txt. Validity blocks GPTBot from its /resources/ section only.
- Three vendors block training crawlers such as GPTBot, ClaudeBot and CCBot at the firewall, even though their robots.txt says Allow.
- Three vendors put a captcha or block page in front of every scripted fetch, a desktop browser string included.
- Nine vendors show no price a crawler can read, because the figures load only in a browser calculator or sit behind a wall.
Where these numbers come from
- 37 email verification vendor domains, all checked on 2026-09-24 from one connection.
- robots.txt was fetched for each domain and read for rules that name an AI crawler.
- Each homepage was requested eight times, once per user agent: a desktop Chrome string, GPTBot, ChatGPT-User, OAI-SearchBot, ClaudeBot, Claude-User, PerplexityBot and CCBot. The HTTP status of each answer was recorded.
- Pricing pages were checked for prices in the served HTML. A firewall result shows what that user agent got from our connection on that day; robots.txt rules are exact.
Blocking by robots.txt is rare and aimed at OpenAI
A robots.txt rule is the public, polite way to turn a crawler away. Only one vendor in the set uses it against answer engines: NeverBounce disallows OAI-SearchBot, ChatGPT-User and OpenAI's Operator from the whole site.
Validity, which sells BriteVerify, blocks GPTBot from its /resources/ section and leaves the rest open. ZeroBounce, Scrubby, Bulk Email Checker and BillionVerify go the other way and name AI crawlers with an explicit Allow.
| Vendor | What robots.txt says to AI crawlers |
|---|---|
| NeverBounce | Disallow the whole site for OAI-SearchBot, ChatGPT-User and Operator |
| Validity (BriteVerify) | GPTBot disallowed from /resources/ only |
| Hunter | No block by user agent; Content-Signal ai-train=no |
| Allegrow | No block by user agent; Content-Signal ai-train=no |
| ZeroBounce | Explicit Allow for GPTBot, OAI-SearchBot and other AI crawlers |
| Scrubby, Bulk Email Checker, BillionVerify, emailverify.io, Bounceless | Explicit Allow for AI crawlers |
| 26 other vendors | No rule naming an AI crawler |
robots.txt read for each domain on 2026-09-24. MyEmailVerifier's robots.txt itself returned a block page, so it is counted under firewall blocks below.
Firewalls block more than robots.txt does
Three vendors say Allow in robots.txt but turn training crawlers away at the firewall. emailverify.io and Bounceless return 403 to GPTBot, ClaudeBot and CCBot. Xverify returns 403 to GPTBot only.
Those three still answered ChatGPT-User, OAI-SearchBot and PerplexityBot with a normal page. So a live answer engine can still read them. The block affects what future models learn from, not what today's search answers can quote.
Three more put a challenge in front of every scripted request. NeverBounce served a PerimeterX captcha, MyEmailVerifier a Cloudflare block page, and EmailHippo a Cloudflare challenge on its main domain and a 403 on its docs.
| Vendor | Firewall result on 2026-09-24 |
|---|---|
| NeverBounce | 403 captcha to every user agent, desktop Chrome included |
| MyEmailVerifier | 403 block page to every user agent, robots.txt included |
| EmailHippo | Challenge on the main domain; docs returned 403 to every user agent |
| emailverify.io | 403 to GPTBot, ClaudeBot and CCBot; 200 to search and user bots |
| Bounceless | 403 to GPTBot, ClaudeBot and CCBot; 200 to OAI-SearchBot and PerplexityBot |
| Xverify | 403 to GPTBot only |
Each homepage requested once per user agent from one connection on 2026-09-24. A 403 here means blocked for that request, which may differ from a residential browser.
Blocked pages still get cited, from other people's copies
Blocking a crawler doesn't take a brand out of AI answers. In 64 search probes run for this study, NeverBounce appeared among the cited sources 8 times and MyEmailVerifier 15 times.
Those answers come from a search index, reviews and other people's pages. When the vendor's own page can't be read, you get what someone else wrote about it. That copy could be years old.
That's why our verifier pricing index shows no figure for a vendor whose own page is unreadable. A price copied from a review can be years out of date.
Nine vendors hide their prices from crawlers
Most crawlers read only the HTML the server sends. A price that loads inside a JavaScript calculator never reaches them. Kickbox, Clearout, Verifalia, QuickEmailVerification, BounceBan, Scrubby, MailerCheck and Listmint price this way. BriteVerify publishes no price at all.
MillionVerifier has no /pricing page and shows its figures on the homepage. NeverBounce and MyEmailVerifier returned 403 on their pricing pages. ZeroBounce keeps its help centre behind a login.
If you're asking an AI assistant what a verifier costs, check the answer against the vendor's own page in your browser. For the vendors whose prices could be read, the pricing index dates every figure.
How to check whether a site blocks GPTBot
Open the site's /robots.txt in your browser. Look for a User-agent: GPTBot line, or OAI-SearchBot, ChatGPT-User, ClaudeBot or PerplexityBot, followed by Disallow: /. That's a full block for that crawler.
A robots.txt that says Allow is not the whole story, as the firewall results above show. To see what a crawler really gets, request the page with its user agent string. Then read the status code: 200 is open, 403 is blocked.
Training crawlers and search crawlers are separate. GPTBot, ClaudeBot and CCBot gather training data. OAI-SearchBot, ChatGPT-User, Claude-User and PerplexityBot fetch pages to answer a question now.
zapbounce.com is open to every answer engine
zapbounce.com's robots.txt allows every crawler, and names GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-User, Claude-SearchBot, PerplexityBot, Perplexity-User, Google-Extended, Applebot-Extended and Bingbot with an explicit Allow.
Every ZapBounce price sits in the page HTML, so any crawler can read it. It's $5 for 1,000 checks, $139 for 100,000 and $949 for a million. Unknowns and duplicates are never billed, and credits never expire. See the full list on the ZapBounce pricing page.
Sources
Questions people ask
Do websites block GPTBot?
Some do, but few in email verification. Of 37 verification vendors checked on 24 September 2026, three blocked GPTBot at the firewall and Validity blocked it from one section in robots.txt. Most had no rule naming GPTBot at all.
Which email verifier blocks ChatGPT?
NeverBounce. Its robots.txt disallows OAI-SearchBot, ChatGPT-User and Operator from the whole site, read on 24 September 2026. It was the only one of 37 verification vendors to block ChatGPT's search crawlers that way.
What is the difference between GPTBot and OAI-SearchBot?
GPTBot collects pages that may be used to train OpenAI's models. OAI-SearchBot and ChatGPT-User fetch pages to answer a question in ChatGPT search right now. Blocking GPTBot alone leaves a site readable in live ChatGPT answers.
Does blocking AI crawlers stop a brand appearing in AI answers?
No. In 64 search probes run for this study, NeverBounce was cited 8 times and MyEmailVerifier 15 times, even though both block scripted fetches. The answers drew on search indexes and third-party pages instead of the vendors' own.
Check your own list's coverage
100 free checks a month, no card. Every result comes back as valid, invalid, catch-all or unknown, and unknowns are never billed.