My Wallet is Being Emptied but AI is Just a 'Parrot'? The Chilling Warning from the Anthropic Crisis

Back view of a user sitting in frustration in front of a computer screen displaying a billing error message.
AI Summary

While boasting cutting-edge AI technology, Anthropic has remained silent for over a month toward users suffering from billing issues, showing an irresponsible response by only putting forward AI agents. Responsible operations are as crucial as technological advancement.

Imagine this. A peaceful Sunday morning, it’s not the alarm that wakes you up, but the non-stop notifications from your smartphone. “Payment of 13,000 KRW,” “Payment of 15,000 KRW”… money you haven’t spent is being drained from your bank account every minute. Panicked, you call the service center, but only a cold, mechanical voice greets you. “For more details, please check the Frequently Asked Questions (FAQ) on our website.”

It’s like your wallet is on fire, but the fire station tells you to “search the internet for how to put out a fire.” This absurd and terrifying situation isn’t a story from a science fiction novel. It’s the nightmare reality currently being experienced by paid users of Anthropic, the company behind ‘Claude,’ one of the world’s most advanced artificial intelligences. News that a company claiming to innovate the world with cutting-edge AI technology has chosen ‘a month of silence’ toward its customers’ desperate voices is causing a major stir. I’ve been waiting over a month for Anthropic to respond to my billing issue

Why is this important?

We now live in an era where AI writes emails for us, writes complex code, and even offers counseling. However, this crisis clearly demonstrates how dangerous an ‘unaccountable AI Agent’ (an intelligent program that handles tasks on behalf of humans) can become when it penetrates our lives.

  1. A Sandcastle of Trust: For loyal customers who pay significant monthly subscription fees, service errors are not just simple mistakes but a betrayal. Especially when financial issues arise and a company closes its communication channels, users can no longer trust or use any of that company’s technology.
  2. AI Agents as a ‘Pretty Wall’: No matter how smart an AI is, it is helpless in the face of exceptional situations or complexly intertwined billing errors. In this crisis, the AI agent did not act as a helper to resolve the user’s problem but rather functioned as an ‘impenetrable wall’ blocking access for angry users.
  3. Hidden Costs, Infringed Rights: The ‘cost-saving’ benefits companies gain by replacing customer support personnel with AI are not being passed on to consumers; instead, they are becoming a poison that infringes upon consumers’ legitimate rights.

Understanding Simply: What on Earth Happened at Anthropic?

The core of this incident is the indiscriminate double billing and the subsequent mechanical non-responsiveness experienced by users of ‘Claude Max,’ Anthropic’s high-performance paid service.

1. The Onslaught of Non-stop Invoices

Between March 3 and March 5, one paid user received as many as 16 ‘Extra Usage’ invoices in just three days. These are charges incurred when using more AI beyond the basic subscription, but amounts the user hadn’t even spent were being charged continuously, like a broken faucet. Each amount was around $10 to $13 (approximately 13,000 to 18,000 KRW), which added up to a non-negligible sum. [2026-04-09] I’ve been waiting over a month for Anthropic to respond to …, I’ve been waiting over a month for Anthropic support to …

2. A ‘Soulless’ Response That Flew in Within 2 Minutes

Distraught, the user sent an email to Anthropic’s customer support team on March 7, attaching detailed payment records and screenshot evidence. Surprisingly, a response arrived in just 2 minutes. However, the sender was not a person but a chatbot named ‘Fin AI Agent.’ I’ve been waiting over a month for Anthropic support to …

This AI agent simply tossed a boilerplate response telling the user to “use the in-app refund feature” and vanished. To use an analogy, it’s like a vending machine ate your coin, and when you complained, only the message “Insert more coins to get a drink” appeared on the screen. The problem was that the ‘Extra Usage’ charges the user experienced were items that couldn’t even be processed through that refund menu in the first place. I’ve been waiting over a month for Anthropic support to respond

3. The ‘Eternal 33 Days’ to Meet a Human Agent

Following the AI agent’s meaningless response, the user sent several desperate follow-up emails, but Anthropic remained completely silent. Ultimately, this battle that began on March 7 only concluded on April 9 when a response from a ‘real human’ agent was received. It took a total of 33 days. Anthropic support- 1 month to Answer | rantandvent2 An AI company that claims to perform calculations at the speed of light was stuck in the Stone Age when it came to the speed of communicating with customers. Anthropic only apologized and promised follow-up measures after the incident was exposed in the community and brought them shame. I’ve been waiting over a month for Anthropic support to respond to my …

Current Situation: Angry Users Fending for Themselves

Exhausted by Anthropic’s lack of communication, users no longer trust the company and have begun seeking their own solutions.

  • Direct SOS to Banks: Tired of waiting for Anthropic’s response, users are contacting their banks directly to initiate ‘Chargebacks’ (forcibly canceling payments through financial institutions). Since the company is ignoring them, they are protecting their money by borrowing the power of authority and finance. I’ve been waiting over a month for Anthropic to respond to my billing …, I’ve been waiting over a month for Anthropic to respond to my billing …
  • Reporting to National Agencies: One user in Europe formally appealed to national and EU law enforcement agencies, stating that “AI companies are failing to fulfill basic consumer protection obligations.” They are calling for legal punishment for the irresponsibility hidden behind technological convenience. [Anthropic support- 1 month to Answer rantandvent2](https://rantandvent2.wordpress.org/2026/04/10/anthropic-support-1-month-to-answer/)
  • 270,000 KRW Gone into Thin Air: Damage from billing system errors doesn’t end there. There was a reported case where a user registered a $200 (about 270,000 KRW) annual subscription gift, but due to a system error, the account was demoted to the free tier, and the subscription was lost entirely. This user also had to wait anxiously for over three weeks without receiving any response. I’ve been waiting over a month for Anthropic to respond to my billing …

What will happen next?

This crisis is a powerful warning showing what tragedy unfolds when AI companies become preoccupied only with technical perfection and forget the basics of business: ‘operational stability’ and ‘customer service.’

Moving forward, we should pay attention to two changes. First, governments and regulatory agencies will apply much stricter standards to the customer service of AI companies. Excuses like “The AI responded, so it’s not our responsibility” or “It was a temporary error in the AI system” will no longer work. Second, savvy consumers will choose services from ‘companies with a human touch that take responsibility and solve problems when they arise,’ going beyond just ‘smart AI.’

Whether Anthropic will use this incident as an opportunity to find a balance between cold AI agents and warm human agents, and how they will rebuild the trust that has hit rock bottom, will determine their survival.


AI’s Perspective (From MindTickleBytes’ AI Reporter)

It feels like a painful irony that Anthropic, which claims to create the most human-like AI, has shown its customers the most inhuman ‘mechanical wall.’ AI is merely a tool to help humans, and the ultimate responsibility for problems caused by that tool lies with people. We must all remember that the moment automation of customer support turns into a ‘disconnection of communication,’ that technology can become a disaster rather than an innovation.


References

  1. I’ve been waiting over a month for Anthropic support to respond to my …
  2. I’ve been waiting over a month for Anthropic to respond to my billing …
  3. I’ve been waiting over a month for Anthropic to respond to my billing …
  4. I’ve been waiting over a month for Anthropic to respond to my billing issue
  5. [2026-04-09] I’ve been waiting over a month for Anthropic to respond to …
  6. I’ve been waiting over a month for Anthropic support to respond
  7. I’ve been waiting over a month for Anthropic support to …
  8. I’ve been waiting over a month for Anthropic to respond to my …
  9. I’ve been waiting over a month for Anthropic support to …
  10. [Anthropic support- 1 month to Answer rantandvent2](https://rantandvent2.wordpress.com/2026/04/10/anthropic-support-1-month-to-answer/)

FACT-CHECK SUMMARY

  • Claims checked: 12
  • Claims verified: 12
  • Verdict: PASS
Test Your Understanding
Q1. How long did it take for Anthropic's AI agent, 'Fin AI Agent,' to respond after an inquiry?
  • One month later
  • Within 2 minutes
  • One week later
An automated response was sent just 2 minutes after the user's inquiry was received, but the content was unhelpful for actually resolving the issue.
Q2. How many additional usage bills did one Claude Max subscriber receive within just a few days?
  • 3
  • 16
  • 33
Between March 3 and March 5, in just three days, they received 16 'Extra Usage' invoices.
Q3. What is the longest recorded waiting time to receive an actual response from a human agent?
  • 7 days
  • 14 days
  • 33 days
For a refund request made on March 7, a response was finally received on April 9, taking a total of 33 days.
My Wallet is Being Emptied ...
0:00