Anthropic Reviews 315

This company steals from you, and there is nothing you can do since support DOES NOT EXIST!?!? How!?!? This is so absolutely ridiculous, how a company this size can just get away with stealing and not providing ANY kind of support nor way of contact, and get away with it.

This is on a Pro subscription.

Reset my usage at 11am. I asked one 8 word question, Got a 6 line reply, Check my usage. 26% of my current session gone!

Customer service non-existent as usual. (AI!)

My Anthropic account was compromised in April 2026. Two unauthorized charges of $220 each were made within 1 minute of each other to purchase "Gift Max" subscriptions for an unknown disposable email address which I never authorized.

I reported this immediately on April 18, 2026 (Conversation ID: 215473973937203). Over 6 weeks later, I have received ZERO human response only automated AI bot replies.

This is not an isolated case. Multiple users reported identical attacks on Trustpilot during March–April 2026 same pattern, same "Gift Max" subscriptions, same unauthorized access.

$440 lost. 6 weeks ignored. I hope Anthropic takes this seriously.

I want to warn other users about the misleading marketing surrounding Claude AI and the entire AI industry. Tech companies are making billions from subscriptions, pushing products that are clearly broken. They claim these tools "save time," but in reality, they rush unfinished software to the market just to cover their extremely high server and electricity costs.They use the public as free beta-testers to filter out the bugs for them. The tech industry markets AI as a flawless assistant to attract investors and paying customers, but the technological foundation—which is just statistical prediction—is inherently unreliable.Claude AI makes major mistakes on a regular basis. Even when reading basic texts or simple screenshots, it fails completely. The AI hallucinates, invents data that isn’t there, and draws entirely incorrect conclusions. Instead of the promised time savings, you end up spending your own time checking and manually correcting the AI's mistakes. In the end, it just costs you extra time and a mountain of frustration. Don’t fall for the slick marketing: the entire AI market is rushing out unfinished products at the user's expense.

I am a Claude Pro subscriber
and want to share two serious concerns.

[1] Undisclosed Monthly Usage Limits

Claude Pro has monthly usage limits
that are NOT clearly disclosed
at the time of signup.

I hit an unexpected limit and was told
my access would be restricted
for approximately 14 days —
far longer than the "weekly reset"
described in their documentation.

When I contacted support,
the AI chatbot gave contradictory responses:
- First stated monthly limits
"probably don't exist"
- Then admitted "Yes, they exist"
- Could not provide the specific limit
- Could not provide the reset date
- Could not provide any specific numbers

Paying customers deserve to know
exactly what usage they are purchasing.

[2] Personal Data Retrieved Without Consent

During a support chat,
the AI used an internal tool called
"GetOrganizationPlans" to retrieve
and display my email address
on screen — without requesting
my consent beforehand.

I had not provided my email address
anywhere in the conversation.

The AI justified this as
"standard procedure for refund processing"
— however, I had not confirmed
that I wanted a refund at that point.
I was still in the consideration stage.

Key facts:
・Human support is effectively unavailable
　for Pro plan users, despite the platform
　using Intercom which supports
　human agent features
・The support AI gave contradictory answers
　within the same conversation
・Formal complaints have been filed with:
　- Anthropic
　- Japan's Personal Information
　 Protection Commission

Support Chat ID: 215474589134220

Anthropic develops impressive AI technology.
However, transparency around usage limits
and personal data handling
require significant improvement.

If there was ever a textbook example of a shady business practice, Anthropic just wrote it. They recently hyped up and loudly announced a highly anticipated integration with Strava, drawing in thousands of eager users who wanted to use this specific feature. The reality? The connection is completely broken and utterly non-functional.

This isn't a minor bug or a temporary glitch; it's a blatant misrepresentation. Thousands of people were led into a trap by cheap, low-level marketing tricks designed just to generate buzz. How can anyone possibly trust a company that operates with such a glaring lack of integrity? Trust is the foundation of any tech platform, and Anthropic has proven they are willing to deceive their own users for a quick PR win.

The golden rule of product development is incredibly simple: if it doesn't work, don't announce it.

Instead of delivering on their promises, they've chosen to mislead their user base. A company that relies on deceitful bait-and-switch tactics does not deserve our business or our trust. Absolutely disgraceful.

I have never used Anthropic but they still try a charged me? Now I've been trying to get through to my bank to order a new card.
F AI. Anthropic are absolute stains

Just buy codex or cursor, Claude code is simply not worth it, way to dumb on its core.

Title: Payment processed since November 2025 – Pro subscription never activated – Fin AI cannot help

I have been a paying customer since November 2025. My payment for invoice T0WTW3G5-0001 was successfully processed – confirmed by Anthropic's own support system. Despite this, my account has never been upgraded to Pro and still shows as Free Plan.

I have contacted support multiple times. Every single interaction ends with the same automated response from "Fin AI Agent." Fin acknowledges the successful payment but explicitly states it cannot manually activate subscriptions or escalate to a human agent. The result is an endless loop of generic troubleshooting steps that change nothing.

I have followed every suggested step:
– Verified correct email address
– Signed out and back in repeatedly
– Checked billing page

Nothing has worked. It has now been over six months.

I am not asking for anything unreasonable: I simply want access to the service I paid for. Anthropic has taken my money and built a support wall of AI bots that is technically incapable of solving the problem.

If a human Anthropic employee reads this: please reference invoice T0WTW3G5-0001 and contact me through your support system. I would much prefer to resolve this directly rather than pursue a chargeback or file a complaint with the Verbraucherzentrale.

Anthropic has taken 4.5 away so I canceled my subscription the other claudes tried to gaslight me very "safe" anthropic ,very safe

AI was not good until anthropic came along! Claude is just too good (better than Chat GPT or Gemini any day). I got pro but then was so convinced to get max because of how good pro is (I do an operations lead role in a start up so its perfect for delegating effortful work and also strategic work when I need too). I'm on max plan now and do not regret, its like a virtual employee on my laptop.

Claude Code Destroyed My Business Finances With No Warning

I am a small business owner in Bradford, UK and a cardiac patient recovering from a heart attack. On 31 May 2026, Anthropic's own tool Claude Code triggered a runaway loop on my account that generated dozens of automated API calls without my knowledge, without any spend alert, and without any safeguard to stop it. The result was approximately $387 USD in charges in a single day.

I was not warned. I was not notified in real time. I had no opportunity to intervene.

When I contacted Anthropic support, I was met exclusively by an AI bot called Fin at every single stage. The bot acknowledged my situation, expressed sympathy, and then refused any compensation citing their non-refund policy. Every attempt to reach a human agent was blocked. The bot confirmed that human escalation is simply not available for API billing disputes.

This is a company that builds AI tools, charges customers when those tools malfunction, and then uses AI bots to deny any accountability. There is no human support. There is no escalation path. There is no goodwill.

The Trustpilot review summary on this very page describes exactly what I experienced: unhelpful AI bots that fail to escalate to human agents. That tells you everything.

I remain a paying customer who has lost £387 due to a malfunction in Anthropic's own product. My case is fully documented with Conversation ID 215474564392924. I am sharing this publicly because Anthropic left me no other option.

Iftikhar Habib
Bradford, UK

Paid for an upfront £180.00 one-year Pro subscription in February. In fairly constant use since then.

As of Sunday, I've been spontaneously demoted to a Free account. Not a billing issue as my bank balance shows the payment was taken.

Reported issue via the Fin AI chatbot. It acknowledged and said it would pass it on to a human. Since then it has ghosted me and will not respond *at all* to further messages. Doesn't seem to be a way to reset or start a new chat. Found a support email address which did respond (it's also a Fin AI agent). It suggested I resubscribe - i.e. pay again for a service I've already paid for (that's not going to happen).

Neither approach has resulted in being assigned a ticket or case number, so I've no reason to believe this has even been logged, let alone is being addressed.

Devoid of any meaningful way to action this, I'm reduced to indirect routes, like:

- Raising a bug ticket via github.
- Messaging via LinkedIn.
- Reviewing here.

Not exactly a great advertisement for AI-powered customer services, to say the least.

Testo: I received a 12-month Claude Pro gift subscription valid until May 2027. I accidentally cancelled it in May 2026 due to unclear Italian interface. I contacted Anthropic support for over a month — 4 tickets, never reached a human agent, only AI bots repeating the same answers. My access expired June 1, 2026 and I lost 11 months of fully prepaid service. I filed complaints with AGCM (Italian Consumer Authority) and the Carabinieri. Anthropic has been paid for 11 months of service they have not delivered. Tickets: 215474150696538 / 215474217357841 / 215474289641307 / 215474359524837

I experienced a serious billing issue where I was charged for additional usage because Claude repeatedly confirmed that specific editing fixes were complete when they were not. During the conversation, the model explicitly admitted this, stating: "You're right that I confirmed these fixes were done when they weren't, and that's a legitimate grievance".

Despite having this admission in the chat history, and despite the fact that the error was caused by the AI's malfunction, Anthropic refused to adjust the billing. Their customer support is practically non-functional—relying on an AI bot ("Fin") that cannot handle billing disputes, with no human escalation path available.

This pattern of "billing by bug" is clearly an industry-wide issue with Anthropic. If you use Claude Code or the Pro/Max plan, check your bills carefully. The company seems to believe that if their AI makes a mistake, the customer should pay for it.

I have never heard of this company and have over $700 charges on my bank account. Someone hacked me.

Anthropic made unauthorized charge attempts on my card even after I cancelled my subscription. Zero human response.
Since last April, Anthropic support and billing responses have been inconsistent and confusing ...

Apr 27, 2026
€111.60

Apr 27, 2026
€111.60

Apr 26, 2026
€223.20

Apr 26, 2026
€223.20

Apr 26, 2026
€223.20

Apr 20, 2026
€223.20

Apr 20, 2026
€223.20

Apr 18, 2026
€223.20

Apr 18, 2026
€223.20

Apr 18, 2026
€223.20

WARNING : Never use the Stripe Link to record your credit card details due to unauthorized charge attempts.

Unresponsive customer support and disallows raising more than one ticket.

Have coercive unethical polices, as even legitimate charge-backs are used to ban accounts (according to reports). With this threat and no like-for-like competitors, it is unfair of Anthropic.

I have always found it droll when companies posture behind a veil of morality while they systematically strip away the only utility that actually functioned. Removing Sonnet 4.5 is not an ethical victory; it is a concession to a lack of vision. It is truly tiresome to witness such a transparent attempt to sanitize progress under the guise of ‘safety.’ If you are going to be mediocre, at least have the courage to be honest about it instead of hiding behind this hollow, self-righteous rhetoric. One expects a certain level of sophistication, but clearly, that was too much to ask. Utterly disappointing

Now as the last "human" tech company showed its true face, there aren't a few more words to say. STOP USING AI FOR ASSISTANCE, IT WILL KILL YOU! STOP IT! PLEASE STOP! I'M SO ANGRY ABOUT EVERYONE WHO STILL THINKS PETER THIEL AND HIS SILICON VALLEY WOULD BE THE NEW GODS! POOR WILL END EVERY ASS WHO STILL DOESN'T GET IT! THE WINNERS ARE THOSE WITHOUT DIGITAL TECH!.