ISPing.me

What to Do When Claude Asks for KYC: Triggers and How to Avoid Them

💡 Key takeaways

KYC isn't for everyone — it's triggered by your environment: datacenter IPs, virtual machines, multi-user sharing, and a card from a different country than the IP all raise the chance of verification. Claude currently does not accept mainland Chinese ID cards or passports, so if verification is triggered and you have no genuine document from another supported region, the account is basically stuck — which is why the real solution comes before the trigger.

A "verify your identity" pop-up at login, asking you to upload documents — before you go hunting for documents, stop. KYC isn't something every user has to do; it's triggered by your environment.

What KYC is: a "triggered" check, not a mandatory step

KYC (Know Your Customer) is how a platform confirms "you're a real person, not a bulk-registered bot account". Claude's identity verification is handled by specialist KYC providers like Persona, and the process itself is standard: upload a document, confirm the details, wait for review.

The key point is this:KYC is not part of the registration process — it's a result of risk controls. A new account with a clean environment may never be asked to verify; a suspicious one can be asked to upload documents right after signing up, even if it has done nothing.

What's worse,Claude currently does not accept mainland Chinese ID cards or passports. For mainland users, that means once verification is triggered without a genuine document from another supported region, the path is basically closed — so the real solution comes before the trigger: don't let your environment attract risk controls in the first place.

How a KYC trigger happens

For mainland users, the sticking point is step three

1Environment anomaly

Datacenter IPs, virtual machines, multi-user sharing, and similar signals are detected

2Risk flag

The account is asked to verify it's a real person (KYC)

3Documents not accepted

Mainland Chinese ID cards and passports are not accepted

4Account stuck

Verification fails and the account is restricted

How a KYC trigger happens

A trigger isn't caused by one signal alone — it's a chain: an environment anomaly is detected first, the account is flagged, then it's asked to verify it's a real person, and finally it gets stuck because the documents aren't accepted and the account is restricted. For mainland users, the sticking point is step three.

  1. Environment anomaly: datacenter IPs, virtual machines, multi-user sharing, and similar signals are detected;
  2. Risk flag: the account is asked to verify it's a real person (KYC);
  3. Documents not accepted: mainland Chinese ID cards and passports are not accepted;
  4. Account stuck: verification fails and the account is restricted.

The first two steps can be acted on; the third has basically no workaround. Risk systems look at the combination of signals, not one isolated fact: a single datacenter IP might only cost a few points, but add a virtual machine, multi-user sharing, and mismatched payment details, and the score climbs. SeeWhy Is a Proxy IP's Risk Score Higher Than a Residential IP's? for why a datacenter IP costs points. That's also why "someone else used the same provider without a problem, but I got flagged the moment I used it" happens — your combinations of environment signals were never the same.

The four environments most likely to trigger KYC

EnvironmentWhy it triggers
VPN / provider nodesThe exit is a datacenter IP, which is exactly what risk systems watch most closely
Using it in a virtual machineNormal users use phones and computers; a virtual machine environment stands out as an abnormal device signal
"Fake residential" IPsMany so-called residential IPs on the market are actually datacenter forwarding — one check of the ASN and PTR gives them away
Genuinely residential but shared by many usersSeveral users on the same IP in the same period — no matter how clean the behavior profile, it gets flagged

What the four have in common: they all make the account look unlike "an ordinary person using it normally at home". "Fake residential" is the easiest to underestimate: the merchant says residential, but when you check checking the ASN reveals the holder is a datacenter; the PTR record doesn't match a residential operator either, and one check gives it away. Before buying, verify first — seeHow to Verify a Proxy Before You Buy It. SeeThe Difference Between Residential and Datacenter IPs for how to tell whether the IP you hold is truly residential or datacenter forwarding. True residential IPs shared by many users are another story: the IP itself is clean, but with several users on it in the same period, the heavy sharing profile still gets flagged. SeeWill Multiple AI Accounts Sharing One IP Get Linked?. The four environments also amplify each other, so check all four.

Payment method is also a risk signal: the card and the IP must be from the same country

Many people focus on the IP and ignore the payment step. If the issuing country of the bank card doesn't match the IP's location at subscription time (say, a US IP with an Asian card), risk systems see a set of contradictory signals. In practice, the safer combination isthe same-country principle: a UK IP with a UK bank card, a US IP with a US bank card.

Another common route is subscribing through App Store or Google Play in-app purchases — but note that the store account's region also has to match the IP, or you'll get stuck at payment all the same.

How KYC triggers differ by situation

Triggered at new-account sign-up. This is the most common kind: the account was just registered and has done nothing, yet verification pops up at login. The cause is usually the environment, not what you did — the profile of a datacenter IP or multi-user sharing was recorded at the very moment of registration. SeeWhy Does Claude Ban Accounts?, and the risk logic at sign-up is the same as for bans.

Triggered at subscription payment. Everything works fine until you bind a card and pay, and then verification pops up. These triggers are mostly about payment signals: a card from a different country than the IP, or a phone number that doesn't match the region, are common causes.

Triggered after switching nodes. It was working fine, then one node switch triggered it. The IP location changed, the ASN changed, and the time zone and language didn't follow — the profile broke apart. SeeDoes It Matter If Your IP Location Doesn't Match Your Real Location?. Before switching, checkpurity.

Is your Claude environment clean? Check it free on ISPing.me right now: Claude Check → — exit region, IP type, the three leaks, and consistency, all in one view. Don't wait until KYC pops up to investigate.

One low-trigger way to use it

Use a genuinely local network for your exit. Roaming data on an overseas mobile plan is a setup many people use — the operator natively assigns a local IP, for one person only, with a genuine location, so the profile is indistinguishable from an ordinary local user. To tell whether it's truly residential, check The Difference Between Residential and Datacenter IPs, and check whether the ASN holder is a local operator.

IP fixed and for your exclusive use. Don't switch nodes randomly and don't carpool-share with others — check the purity and type before using. For the difference between a fixed IP and a dynamic IP, seeWhat's the Difference Between a Static IP and a Dynamic IP?. Purity = 100 − risk score, and the risk score has five tiers: below 20 pure, below 35 fairly pure, below 50 average, below 70 suspicious, and 70 or above high-risk. For how to lower a high risk score, seeWhat to Do About a High IP Risk Score? How to Lower It. Note that this is only a heuristic estimate based on public information, not an official ban decision from any platform; residential and datacenter IPs have different clean baselines, so comparing numbers across types is meaningless. SeeWhat Is IP Purity? How to Read the Risk Score for what the five risk tiers mean and how to read them.

Fixed device, same-country payment. Use one phone or computer as much as possible, no virtual machines, and don't hop between devices; keep the bank card's issuing country consistent with the IP's location. Time zone and language are bonus points, ranked below IP and payment.

What to do if you've already been asked for KYC

If you have a genuine document from a supported region, verify through the process — that's the only proper route. Upload your documents as Persona instructs, confirm the details, and wait patiently for review. Don't switch IPs, devices, or browsers midway.

Don't photoshop documents or buy fake ones. Verification providers are specialists — image forgery basically never fools them. Even if you get through once, a later risk review will still disable the account.

Without a valid document, the account probably can't be saved. Rather than struggling repeatedly, put your energy into the new account's environment — don't make the same mistake twice after being triggered once. Before registering a new account, check the environment first: exit region, IP type and consistency, WebRTC, DNS, and IPv6 leaks — only proceed when all 9 pass. For how to check, seeCheck These Three Things Before Using Claude.

Phone numbers and virtual cards: if the IP doesn't match, top-ups get rejected too

The phone number and card are only half of KYC — the other half is the environment: if the phone number's country, the virtual card's issuing country, and the IP's location don't line up, top-ups are easily rejected or asked for a second verification. Use the same fixed line from the country as the number and card when you get the card and the number; don't switch nodes at the moment of payment. Before paying, use the disguise check to confirm that fingerprint, time zone, language, and IP all point to the same region — fill any gap before paying.

FAQ

Does every Claude user have to do KYC?
No. KYC is a risk-control trigger, not part of sign-up. A clean account may never be asked to verify, while a suspicious one can be asked for documents right after registering, even if it has done nothing. It comes down to the combination of environment signals, not the account's age.

Can mainland Chinese ID cards really not be used at all?
Under the current policy, Claude does not accept mainland Chinese ID cards or passports. Policies can change — go by the list of accepted documents published on the official website.

If I switch to a clean IP, will a flagged account recover on its own?
No. Triggered verification doesn't disappear if you change IPs. With a valid document from a supported region, verify through the process; without one, the account is most likely unrecoverable. Instead of switching IPs repeatedly, focus on the new account's environment — don't repeat that mistake.

If my IP is clean, will KYC definitely never be triggered?
No. The IP is only one signal — virtual machines, a card from another country, or a mismatched phone number can trigger it too. The whole environment has to be consistent: IP, device, payment, time zone and language. A gap in any one of them can get you flagged — watching the IP alone isn't enough.

Notes

Risk strategies change over time; this article describes the trigger logic that's common today, not a restatement of Claude's official policy. The accepted document range and verification process follow what's published on the official website. Purity and risk scores are heuristic estimates based on public information — treat the results as a reference only; how the same IP performs on different platforms still depends on each platform's risk controls. Don't buy fake documents and don't borrow someone else's — those actions alone can get an account permanently banned.

Check your Claude environment

Enter your IP on ISPing.me to see the exit region, IP type, purity, and consistency results — only register and pay when all 9 pass

Check now

Related reading:Why Does Claude Ban Accounts? IP and Environment Consistency ExplainedWhat to Do When Claude Says Your Region Isn't Supported?Check These Three Things Before Using Claude: WebRTC, DNS, and IPv6 LeaksWhat Is IP Purity? How to Read the Risk ScoreResidential vs. Datacenter IPs: What's the DifferenceWhy Does Dual ISP Still Show as an "IDC Datacenter IP"?Will Multiple AI Accounts Sharing One IP Get Linked?

IP Knowledge Base