Skip to content
Claudexia TeamERRORS

Your account can't receive replies from Claude: what to do

What Claude's Anthropic Verification Portal and paused-organization messages mean, what Anthropic's help pages say to do, and what not to do while you wait.

"Your account can't receive replies from Claude right now" means that Claude's replies are unavailable for your account, and Anthropic points you to the Verification Portal for the next step. The message does not say what triggered the check. Follow the help pages: open the portal through the link in the message, and if the notice is about an organization, request a review of it.

Message texts and what to do

These are the texts people bring to search. Rows one and two quote the English strings exactly. Rows three and four describe notices whose exact English wording we did not find.

Message textWhat is knownWhat to do
API Error: Your account can't receive replies from Claude right now. You can find the next step in the Anthropic Verification Portal: followed by a linkThe same text with a link shaped like portal.anthropic.com/verify?org=... appears in the claude-mem tracker (issue 5251, 1 November 2026, Claude Code 2.1.273, subscription login) and in the okou tracker (issue 44146, records from 29 October 2026). Anthropic's help pages do not analyze itOpen the portal through the link in the message and follow the steps it shows
Replies from Claude are unavailable on your account. You can complete the verification now in the Anthropic Verification Portal.Readers bring this variant to search, and we found no such text in Anthropic's help. The word "verification" suggests the portal expects a check from you, but the text does not say which oneSame as the first row
A notice that an organization you belong to is paused because of unusual activity, with an option to request a reviewAnthropic's help article describes this situation: your own account is fine, but an organization you belong to was paused because of unusual activity. We did not find the exact English wording of that screenClick "Request a review" on the affected organization
A short notice that access to Claude is paused, with nothing after itA headline alone does not tell you whether the pause is on an organization or on your personal accountRead the whole screen: if it lists organizations and offers a review, see the previous row; if your account was banned, you need an appeal (below)

Do not let the portal's name confuse you. Anthropic's documentation mentions the Verification Portal in other programs too, the Cyber Verification Program for one, and puts it at portal.anthropic.com. Your message arrives with its own link, which, judging by user reports, carries an organization identifier.

One more detail from the claude-mem tracker: there the message came back as an ordinary assistant text reply with an API Error prefix. The tool took it for ordinary model output, asked for the batch once more, and after a second such reply dropped it. The reporter counted about 770 batches lost over roughly 25 hours. The fix taught the tool to recognize this reply and keep the batch in the queue (the issue was closed on 2 November 2026). If you build automation around Claude Code, plan for this reply.

What to do, per Anthropic's help pages

The list below follows the Anthropic help pages we opened on 13 November 2026: warnings and appeals, identity verification, and contacting support.

If the message sends you to the Verification Portal

  1. Open the portal only through the link in the message itself, or by typing portal.anthropic.com yourself. If the link in the message points to another domain, do not follow it. After that, follow only links that the portal itself or Anthropic's help pages give you, such as the help form in step 4.
  2. Follow the steps the portal shows. What it will ask in your case is not visible from the help pages.
  3. If identity verification is requested, have the original government-issued photo ID in hand and a phone with a camera: you may be asked for a live selfie. The help page accepts a passport, a driver's license or state ID card, and a national identity card. It does not accept photocopies, screenshots, scans or photos of a photo, digital or mobile IDs, non-government IDs such as student cards or employee badges, or temporary paper IDs.
  4. If verification fails, the help page suggests three steps: retry (there are several attempts inside the flow, and a photo in better light or a different government ID usually fixes it), check that the document is undamaged and legible, and write through the help form.

On data, the page says the ID and selfie are collected and held by Persona, not on Anthropic's systems. Anthropic can open the records through Persona's platform, for example to review an appeal, and does not use the data to train its models. The page names no retention period.

If the notice is about an organization

The help article on paused organizations is short. When you log in, you are shown that an organization was paused because of unusual activity, and the restricted account screen lists such organizations. Click "Request a review" on the affected one and the Safeguards team takes another look. The article gives no further steps.

If your account is banned

Log in to claude.ai with the banned account and fill out the "Submit an appeal" form. It is not available without logging in. The help article names the grounds for bans: repeated Usage Policy violations, account creation from an unsupported location, and Terms of Service violations. You can still export your data or delete the account after a ban, on the same screen after logging in.

Keep to a single conversation with support: the contact guide asks for one conversation per issue, because a second request has to be triaged again and usually slows the answer. The reply appears in the support messenger and arrives by email at the address on the account.

Timing

Anthropic names no deadline. The help article on appeals carries a note that replies are currently slower than usual because of a recent launch and a larger volume of email, with no figures. The support guide says only that response times vary by plan and by the severity of the issue. Any "usually a couple of days" comes from other people's stories, not from documentation.

What Meduza reported

Meduza reported on 29 October 2026 that Claude users on Reddit, X and Telegram say their accounts were banned after identity verification with Russian documents. The check ran through the service Persona: pick the region of the document, upload a photo of it, complete a face scan. Why people were sent to verification, the outlet says, is unclear.

What follows are complaints Meduza relays from users.

  • The ban arrives right after a Russian passport is uploaded, including for people who live outside Russia but have no residence permit in another country.
  • A Russian citizen living in New Zealand on a four-year visa said on Reddit that his account was blocked without explanation soon after he uploaded his passport, two days after a subscription payment was charged.
  • Users with Russian documents in Georgia and Armenia described similar problems. One of them published a letter from Anthropic's security team: automated systems found signals of Usage Policy violations, the team reviewed the findings and revoked access. The letter does not say which violation.
  • Another user could no longer send requests after uploading his passport and could only export his chat data. An appeal did not restore the account.
  • Users with Chinese citizenship faced similar checks, and some of them passed.

The outlet adds that Anthropic had not commented on the complaints, and that the company's policy mentions no restrictions by nationality or citizenship and only lists the countries where its products are officially supported. In early October, Meduza recalls, there were already reports of blocks for users from unsupported regions who were getting around the restrictions with a VPN.

This is a retelling of complaints, with no confirmation from Anthropic. Whether your message is connected to this story we do not know: the Verification Portal text says nothing about a passport or citizenship.

What not to do

This relies on the Usage Policy, which its page marks as effective from 12 November 2026, section Do Not Abuse Our Platform, and on the Supported Regions Policy with the same date. What changed in those documents on 12 November is covered in Anthropic Supported Regions Policy, Nov 12: what changed.

Do not open a new account to replace a banned one. The policy treats an attempt to "circumvent a ban through the use of a different account" as abuse of the platform. The wording covers creating a new account, using an existing one, and giving access to a person who was previously banned. A separate item forbids creating multiple accounts to avoid detection or to get around product guardrails, and another covers automation in account creation.

Do not circumvent sign-up or login. The policy forbids intentionally bypassing the restrictions built into the products, including the sign-up and login process, without Anthropic's prior authorization.

Do not spoof your region. The Supported Regions Policy allows use only in the listed countries and treats everything not on the list as unsupported. The rule covers a person's physical location and entities: those incorporated or headquartered in an unsupported region, and those controlled from there. As of 13 November 2026, Russia, Belarus and China are not on the list on that page. The Usage Policy forbids providing access in violation of the Supported Regions Policy. We do not recommend spoofing your region through a VPN.

Do not upload someone else's or forged documents. In its section on fraud, the Usage Policy forbids creating, buying and selling falsified documents, fake IDs included. The verification help page accepts only the original of your own document.

Do not buy other people's accounts. We found no item about buying specifically. Next to it stand a ban on reselling or proxying access to Claude through unauthorized means and a ban on giving access to a person who was banned. You know nothing about a stranger account's history: that is our own reasoning, not Anthropic's words.

We do not give legal assessments and do not promise that any method is lawful or safe for your account. You read and accept the vendor's terms. Whether verification will pass and access will return, nobody promises, and neither do we.

Keeping your work going during the check

Save what you can first. If the account is banned, the help article allows a data export after login, though what you can export may be limited.

If your work runs through the API and cannot wait, a Claudexia key is issued without an Anthropic account. On the sign-up page, create a login and a password with confirmation, or sign in with Google, GitHub or Telegram: there is no email field. The minimum top-up is $1, payable through SBP (rubles), a Russian bank card or crypto. The address for the Anthropic SDK and Claude Code is https://api.claudexia.tech, and keys start with sk_cdx_. Models and prices are on the models page. The key does not lift a ban and does not replace working things out with Anthropic. It only keeps your tasks from stopping.

The step-by-step setup is in the guides on using the Claude API without an Anthropic account and on getting an API key. If API requests started failing with a 400 after you moved to a new Anthropic API version, that is a different cause, covered in the article on the 400 error from temperature, top_p and top_k.

Neighboring situations have their own guides: “Your account has been suspended”, blocked and banned accounts, and VPN bans and what replaces them.

FAQ

How long will a verification or an appeal take?

Anthropic names no deadline. The help article only warns that replies are slower than usual right now. We do not quote timelines from other people's stories because there is nothing to confirm them.

Can I open a new account while I wait?

The Usage Policy treats getting around a ban through another account, including a newly created one, as abuse of the platform. We do not recommend it.

The message appeared in Claude Code. Is it the same thing?

By the report in the claude-mem tracker, yes: there the text arrived in Claude Code with a subscription login. For logins with a Console API key, Anthropic says nothing about this message, and we did not check.

Is it safe to send a passport for verification?

We cannot assess that for you. The help page says Persona holds the ID, not Anthropic, that the data is not used for model training, and it names no retention period. Meduza reports complaints about bans after passport uploads, with no confirmation from Anthropic. The decision is yours.

Does it have to be a passport?

The help page accepts a passport, a driver's license or a national identity card. Which document the portal asks for in your case is not visible from the help pages.