Some Claude users say Anthropic banned their accounts right after they completed an identity check with a Russian passport, including people who live in countries where Claude is officially available. Meduza collected the complaints on October 29 from posts on Reddit, X and Telegram. Anthropic has not commented on them, according to Meduza, so the cause of the bans is not confirmed.
The reports matter beyond Russia. They show what can happen when Anthropic's identity verification meets its regional rules, and how little an affected user can learn about why access was cut.
What users say happened
According to Meduza, Claude asked these users to verify their identity through Persona, an outside verification company. The users said they do not know why they were asked. Persona has them choose the region of their document, upload a photo of it and complete a facial scan.
Users told Meduza the ban came right after they uploaded a Russian passport. One Reddit user said he is a Russian citizen living in New Zealand on a four-year visa. He said he paid for Claude, used it for work and personal tasks, and also used Claude Code. His account was blocked without explanation right after he uploaded his Russian international passport, he said, two days after he was charged for his subscription.
Commenters with Russian documents living in Georgia and Armenia described similar problems, Meduza reported. Meduza also reported that the bans hit people who do not live in Russia but have no residency permit in another country. A user in Georgia posted a letter he said came from Anthropic's security team. It read in part: "Our automated systems detected a large number of signals associated with your account that violate our Usage Policy." The letter did not say which rule he broke. Another user said he could only export his chats after the ban, and that his appeal failed.
Meduza said some users with Chinese passports faced similar checks, though some of them passed. Russian tech outlet vc.ru separately collected posts from users who said they were banned soon after verifying with Russian or Chinese passports, and one who said Anthropic refunded the money after the ban.
What Anthropic's rules say
Anthropic's Supported Regions Policy lists the countries where its apps, API and cloud offerings are available. Russia and China are not on the list. New Zealand, Georgia and Armenia are.
The policy is written around location and corporate control, not citizenship. It says unsupported use includes use by people "while physically located in an unsupported region," use by entities incorporated or headquartered in an unsupported region, and use by entities majority-owned or controlled from one. Meduza made the same point: the rules do not mention nationality or citizenship.
The corporate rule dates back to a September 2025 post, in which Anthropic said it would bar companies more than 50% owned, directly or indirectly, by companies headquartered in unsupported regions such as China. That post was about companies. It did not describe checks on individual passport holders.
How verification and appeals work
Anthropic's identity verification page says the company is "rolling out identity and business verification for a few use cases." A prompt can appear when someone accesses certain capabilities, as part of routine platform integrity checks, or for other safety and compliance reasons. Accepted documents are original government-issued photo IDs "from most countries," such as a passport, driver's license or national ID card. A live selfie may be required.
The page says Persona collects and holds the ID and selfie, and Anthropic can access the records through Persona when needed, for example to review an appeal. It also answers the question users are now asking. On why an account was banned after verification, it says Anthropic "may ban an account for a variety of reasons, including repeated violations of our Usage Policy or Terms of Service," and points to an appeal form.
The Safeguards warnings and appeals page lists creating an account from an unsupported location as one reason for a ban. Banned users must log in to reach the appeal form. They can delete their account, and self-serve exports are open to all banned users, though the data they can export "may be restricted depending upon the Usage Policy violation associated with your account." The page also warns that response times are longer than normal.
Why it matters
Anthropic uses automated signals to police use of Claude. Its self-assessment monitor is one recent example. The ban letter quoted by Meduza fits that pattern: automated systems flag an account, and a team reviews the flag.
What is not clear is whether a passport is itself one of those signals. If it is, the practical rule for some users is stricter than the written one, since the policy says nothing about citizenship. If it is not, the timing users describe may come from other signals such as location or payment details, which Anthropic does not list. Either way, users who live legally in a supported country and are banned have little to go on beyond a general letter and an appeal form.
Until Anthropic explains the bans, affected users can file an appeal and export their data from the ban screen. Anyone asked to verify should also be careful with unexpected messages that use Anthropic's name, since attackers have used fake Anthropic invites to steal logins.
Sources
- Meduza: Claude users say their accounts were banned after verification showed a Russian passport
- vc.ru: Claude users from Russia and China blocked after passport checks (Russian)
- Anthropic: Supported Regions Policy
- Anthropic: Updating restrictions of sales to unsupported regions (September 2025)
- Claude Help Center: Identity verification on Claude
- Claude Help Center: Safeguards warnings and appeals
