Anthropic’s published rules summarized first, then in-depth guides on risk mechanisms, environment cleanup, payment safety, API safety, and appeal SOPs.
This knowledge base collects what is publicly known about how Claude Code fingerprints users and how bans happen, from Unicode steganography and Anthropic’s multi-layer risk model to environment cleanup, account and payment safety, and what to do once an account is flagged. Statements about Anthropic’s rules link to official pages and carry a verification date. Everything else comes from community reports and public reverse-engineering and is labeled as such. Every article is bilingual and can be read on its own or as part of a path below.
This knowledge base is educational and for reference — not official Anthropic documentation.
What Anthropic states publicly. Last checked 2026-10-08. Everything else in these guides is community experience or reverse-engineering and is labeled as such.
Claude.ai and the API are offered only in the countries on the Supported Regions list. Mainland China, Hong Kong, and Macau are not on it; Taiwan, Singapore, Japan, and India are. Anthropic also reserves the right to refuse entities whose majority ownership is attributable to nations not on the list.
Anthropic lists three: repeated Usage Policy violations, account creation from an unsupported location, and Terms of Service violations.
Source: Claude Help Center: Safeguards warnings and appeals ↗ · Anthropic: Usage Policy ↗
Sign in to claude.ai with the banned account to reach the appeal form. Free, Pro, and Max accounts banned for Usage Policy violations can still export data or delete the account from the same screen. Warnings on API accounts go to usersafety@anthropic.com.
Source: Claude Help Center: Safeguards warnings and appeals ↗
Pro and Max subscriptions bought on the web accept credit or debit cards only; PayPal and Venmo are not accepted. App Store and Google Play subscriptions use the store’s payment methods. The card’s billing address must match its country of origin and be an eligible billing location.
Source: Claude Help Center: Paid plan billing FAQs ↗ · Claude Help Center: Why was my card declined? ↗
For the first half of 2026 Anthropic reports 11.4 million accounts disabled, 398,000 appeals, and 42,000 reinstatements, about 10.5% of appeals.
Source: Anthropic: Transparency Hub ↗
Not sure which article to open first? Pick the path that matches your situation.
Need the essentials fast? Account safety, API limits, and frequency rules in one short guide.
Open the checklist →New here? Start with how the fingerprinting and 4-layer risk model actually work before you change anything.
Read the risk model →Sync your timezone, isolate Chinese fonts, stop WebRTC leaks and set up a residential IP the right way.
Clean up your setup →Choose a verification number, use a card type Claude accepts and align your billing details with the card before you subscribe.
Register the safe way →Keep the CLI and browser consistent, choose relays that preserve prompt caching, and stay inside rate limits.
Configure Claude Code →Already flagged? Diagnose the ban type, send a fact-based English appeal, or fail over to an alternative model.
Start recovery →Official ban reasons and appeal routes first, then a short checklist for account and payment safety, environment hygiene, API limits, and usage patterns, with links to the in-depth guides.
Read Guide →How to choose phone verification, understand virtual-card BIN risk, keep the billing address consistent with the card, and avoid linked bans after a failed renewal.
Read Guide →Complete multi-account isolation strategies including physical vs software separation, browser profile setup, IP/payment distribution, and safe switching procedures.
Read Guide →Which payment methods Claude accepts, how card type and billing address affect approval, what the official decline checklist says, and how to avoid unreliable card sellers.
Read Guide →Region-specific Claude access guide covering supported vs restricted regions, US/UK/SG registration best practices, GDPR compliance, and cross-border travel scenarios.
Read Guide →Deep dive into how Claude Code embedded hidden Unicode markers into system prompts and how Anthropic flags accounts based on IP, timezone, BIN, and usage patterns.
Read Guide →Unverified ban case write-ups covering timezone mismatches, prepaid-card refunds, API abuse, and account association, with prevention checklists. Root causes are interpretations, not Anthropic findings.
Read Guide →Step-by-step guide to syncing OS timezones, isolating Chinese font fingerprints, preventing WebRTC/DNS leaks, and setting up anti-detect browsers.
Read Guide →Practical configuration guide for Chrome, Firefox, and Edge browsers including timezone sync, language settings, WebRTC controls, and essential privacy extensions.
Read Guide →In-depth comparison of residential IPs vs datacenter IPs for Claude access, proxy protocol selection, VPN provider evaluation criteria, and DNS leak prevention.
Read Guide →Platform-specific anti-ban configuration guides for macOS, Windows, Linux, iOS, and Android including timezone settings, environment variables, and font isolation.
Read Guide →How to safely set custom ANTHROPIC_BASE_URL, override TZ env variables, choose high-availability API gateways, and preserve prompt caching.
Read Guide →Advanced techniques for maximizing Prompt Cache efficiency, optimizing token billing, managing long context windows, and implementing robust retry logic.
Read Guide →Enterprise-grade automation safety practices including rate limiting, anti-abuse detection avoidance, multi-account load balancing, and compliance audit logging.
Read Guide →Identify 403 / Refunded / Disabled / organization-hold account statuses, plus a fact-based English appeal email template, data export notes, and refund steps after a ban.
Read Guide →Systematic troubleshooting guide for Claude login failures, API call exceptions, payment binding issues, with diagnostic flowcharts and quick fixes.
Read Guide →Complete guide to DeepSeek R1/V3, GLM-4, and Kimi integrations, plus setting up Ollama and One-API as local Claude API drop-in replacements.
Read Guide →Recent region, policy, and enforcement changes that affect accounts.
To reduce data-exfiltration risk, `web_fetch` in Claude Managed Agents fetches only URLs that already appeared in the session: in a user message, in a `web_search` result, or in a page fetched earlier. A URL that appears only in Claude's own output, the system prompt, an attached document, or tool output such as `bash`, `read`, or an MCP tool returns a `url_not_in_prior_context` error. Separately, in cloud environments with `limited` networking, `allowed_hosts` now also applies to `web_search` and `web_fetch`. Creating or updating a session fails with a 400 error when an enabled web tool's `allowed_domains` has an entry outside `allowed_hosts`. An `allowed_hosts` entry matches one exact host unless it starts with `*.`.
Anthropic merged Project Glasswing and the Cyber Verification Program (CVP) into one program with three tiers. Defense Access covers SOC and incident-response work, malware reverse engineering, and vulnerability analysis; Anthropic aims to respond within a few days. Red Team Access adds authorized penetration testing and red-teaming for organizations only, with a review of a few weeks; applicants are enrolled in Defense Access in the meantime. Specialized Access has the fewest blocks and is reserved for verified organizations authorized to test systems such as power grids and interbank transfer infrastructure, reviewed together with the US government. Enrolled organizations must allow data retention; Enterprise Frontier Safeguards, planned for later this fall, will allow storage in customer-controlled cloud. Generally available models keep conservative cyber safeguards and still handle code review, patching known issues, and alert triage. Existing CVP members keep their settings and are evaluated automatically for the new models.
From October 1, 2026, Hong Kong users reported "Account Suspended" or "Access Denied" screens on claude.ai, including paid Pro subscribers who had connected through VPN exits in supported regions such as Singapore. Hong Kong, mainland China, and Macau are not on Anthropic's Supported Regions list; Taiwan, Singapore, and Japan are. The Help Center lists "account creation from an unsupported location" as a ban reason, and a VPN exit in a supported country does not change eligibility. One posted notice cited suspicious signals and a Usage Policy violation without naming a trigger. Media and community analysis points to shared data-center VPN IPs, virtual-card or unsupported-region payments, and frequent country switching, but Anthropic has not confirmed any of this and the number of affected accounts is unknown. One user reported that a data export returned account information without conversation history; the Help Center says exportable data can be restricted depending on the violation.