I rely on Claude Pro daily for professional engineering and backend development. On October 20, Anthropic gladly processed a commercial subscription payment (€22.50 via Mastercard, including 25% Swedish VAT, with traffic routing through a dedicated node in Stockholm).

On October 30 — right in the middle of a business-critical deadline for my engineering deliverables — my access was abruptly terminated following an automated Persona identity check.

Within 4 hours, Anthropic’s automated compliance pipeline generated three mutually contradictory policy excuses:

First, an alleged Supported Countries Policy issue (triggered solely by the issuing country of my national passport).
Second, their automated Fin AI Agent cited an unspecified Consumer Terms of Service clause specifically to deny a refund for unrendered services, followed by: "This conversation has been closed and is no longer monitored."
Third, their Trust & Safety determination abruptly shifted the rationale to an alleged Usage Policy (AUP) violation — implicitly slandering an ordinary open-source engineer under categories reserved for cyberattacks and illicit activity without providing a single operational log.

Follow-up communications routed to an appeals portal already marked as "Final decision", followed shortly by an automated CSAT survey asking me to "rate my customer experience".

A formal chargeback is in progress via Mastercard, and an official regulatory complaint under DSA Article 17 has been lodged with the Irish Digital Services Coordinator (Coimisiún na Meán, Case: CAS-13677) and Swedish PTS.

Below is my formal Open Letter to Anthropic's leadership.

An Open Letter to Dario and Daniela Amodei: On Code, Arbitrary Borders, and What It Means to Build AI for Humanity

To: Dario Amodei (CEO), Daniela Amodei (President), Jack Clark (Head of Policy)

Company: Anthropic PBC / Anthropic Ireland, Limited

From: Konstantin (Software Developer & Open Source Contributor)

Date: October 30, 2026

Reference ID: TS-01a18cb2-0119-7712-b3bd-a7717134a6d4

Invoice / Receipt: 9BF0758D-9606630 / #2711-4150

Dear Dario and Daniela,

I am writing this not to argue about billing terms, nor to threaten legal procedures. I am writing to you directly, founder to engineer, because Anthropic was founded on a noble premise: to build artificial intelligence that genuinely benefits all of humanity.

Today, I found myself asking a very painful question: Who is allowed to be part of that humanity?

I am a software engineer. My life’s work is written in open-source repositories, backend systems, and lines of code designed to solve problems and create value for developers and users across the globe. Code does not care about geopolitics. A compiler does not ask what language you speak at home, and an algorithm does not care what color your passport is. To me, technology has always been the ultimate equalizer — the one place on Earth where merit, dedication, and the desire to build something useful were all that mattered.

Like so many engineers of my generation, I did not choose where I was born. I would gladly pack my life into a suitcase tomorrow, establish permanent residency in Europe, pay my taxes, and build technology in an environment free from geopolitical crossfire. But borders, visas, and international politics are crushing, complex realities that an individual software developer cannot simply dissolve overnight.

Despite this, I wanted to participate in the global digital economy honestly. I set up my own dedicated server infrastructure in Sweden, routed my work legitimately through Stockholm, used my own hard-earned money to subscribe to Claude Pro, and gladly paid European VAT (Receipt #2711-4150, €22.50). I didn't pirate your software. I didn't hack your APIs. I didn't generate malware or harmful content. I came to Claude as a fellow craftsman — to build, to write code, and to meet a critical deadline for an engineering project.

What happened next broke something inside me.

Because your automated system demanded an identity check, I presented the only official document I possess: my passport. Within minutes, an unfeeling algorithm reduced my entire identity, my work, and my good faith into a single derogatory flag:

Mid-Cycle Revocation: My account was permanently terminated mid-cycle on the exact day of a critical project deadline.
Three Conflicting Reasons in 4 Hours: Your automated systems cited three mutually contradictory policies within hours: first “Supported countries policy”, then a blanket “Terms of Service” clause to refuse a refund, and finally a fabricated “Usage Policy (AUP)” violation — implicitly accusing a legitimate engineer of cyberattacks or dangerous conduct without a single log.
Automated Stonewalling: Your Fin AI Agent stated "They work through these in order, so it won't be today", followed by "This conversation has been closed and is no longer monitored."
The Kafkaesque Loop: Your automated support email sent me a link to an appeals form that your Safeguards team had ALREADY closed with "This is our final decision."
The Insult: An hour later, your system automated an email asking me to "rate my customer experience".

The Deepest Irony: The Training Data Paradox

There is an even deeper, bitter irony at the heart of this ordeal.

Claude was trained on public internet data, open-source repositories, and millions of code commits written by engineers from every corner of this planet. When your web scrapers collected our public libraries, our StackOverflow answers, and our collective architectural patterns, you did not inspect our passports. You did not ask where the authors of the software libraries that power modern computing were born. You took the fruits of global open-source labor freely, without regional discrimination.

Yet today, when an open-source builder who contributed to that shared foundation seeks to use the intelligence distilled from our collective work, you demand papers, deploy biometric filters, and revoke access mid-day.

We gave our work freely to the world. You commodified it, locked it behind arbitrary borders, and turned AI into a gated garden for the geopolitically fortunate.

Dario, Daniela — is this truly the world you set out to create?

When you left OpenAI to build an ethical AI lab, was the vision really to construct digital Berlin walls that punish independent creators for the sins of governments they have no control over? Is it truly "safe and harmless AI" to eagerly take a developer’s payment, and then lock them out mid-month without recourse, keeping their money behind a wall of unreachable bots?

I don't expect you to violate international laws. But there is a profound difference between legal compliance and cold, indifferent corporate dehumanization. Sanctions are meant to target power and aggression; they were never meant to strip independent open-source developers of their tools and dignity.

I don't care about the $25. What hurts is the realization that in the eyes of the company building the "future of human intelligence", I am not an engineer who wants to contribute to the world. I am merely an unwanted variable to be purged by a script.

I will continue writing code, and I will find a way to finish my projects, with or without Claude. But as you continue to build systems meant to guide the future of our civilization, I hope you remember that behind every prompt, behind every IP address, and behind every banned account, there is an actual human being who simply wanted to build something good.

Sincerely,

Konstantin

Software Developer & Open Source Contributor

amadeusbackit@gmail.com

Archived case documentation on GitHub Gist:

https://gist.github.com/markankamaransheogarat-sys/42799d8c9cd54300e57d315f24a16a42

Original communication trail on X:

https://x.com/MarkanKamaranSG/status/2116157295990296873