Anthropic Apposite Appeal

1st and 2nd appeal for intervention by law enforcement …

The 1st (last email in the chain below) is the official complaint making claims with not much evidence, if any…

The first below is my follow up on 7 March ‘26.

This is when Anthropic throttled me back to;

Availability: 5% - 22 hrs 48 min downtime per day

09 March’26

cc: Italian law enforcement (AGCM & Garante) and EU law enforcement (CNECT-AIOFFICE).

Il giorno Mon 9 Mar 2026 alle 14:40 0808@pec.it ha scritto:

Dear sir madam

I am writing to follow up on my earlier mail regarding unfair commercial practices.
Prior to invoking the ad hoc mechanism for increasing the number of ‘tokens’, a typical day’s professional technical session involved:

Ghost migration, SMTP configuration, newsletter setup, editorial work, Reddit strategy, two published essays, delivered in one day. All of the config work involved endless screenshots back to Claude, because my techie days are way behind me. 

I was handheld by Claude every step of the way. If you’ve ever created a web site, then been forced to migrate to a virtual server whilst trying to keep output high, you’ll know there are hundreds of steps involved, each into territory I’ve never seen in my life before, despite being a data network specialist in my past.

At the peak of this intensity of activity is when I reached my subscription limits and sought relief with the ‘wet finger in the wind’ offer of undefined ‘tokens’.  Very bad idea.

I include this account, and forgive me, but it is to provide a contrast against my current level, post subscription cancellation, as well as card blocking to prevent Anthropic from continuing to charge post the cancellation, which they managed twice before I was able to stop it.

At present, I’m constantly being throttled by misanthropic. 5 to 10 minutes is apparently exceeding my subscription limits. Since 04:30 this morning I’ve only been allowed to upload one image.

I’m hitting every limit they can impose. 

This is a deliberate act of retribution for cancelling, blocking and, if they’ve heard from yourselves, revenge for reporting them. If they haven’t heard from you, I can only imagine how much more they’ll squeeze my bandwidth.

I am not making evidence-free claims, Claude has all our conversations in compressed format which makes them difficult to access directly. 

If Anthropic wish to challenge my claims of severe service degradations since cancellation and payment blocking, I invite them to bring the receipts for due diligence.

At the point of cancellation I had 13 days of my subscription left, which should have returned to the pre undefined ‘tokens’ of infrastructure cost being passed to me. 

This has not happened. In fact the opposite as I’m sure you can observe from the contrast I provide above.

I am not aware of your response times, but I would like to recover my paid for, as per contract, subscription without service degradation until the 20th of March 2026 - midnight? They don’t say.

I have every intention of restoring my subscription but not on the present terms. 

I seek your assistance in a return to the days of pre undefined ‘tokens’ of infrastructure cost being passed to me.  

I look forward to your response.

Thanks and regards

0808

Il giorno Sun 8 Mar 2026 alle 08:40 0808@pec.it ha scritto:

xxx, Italy
08 March 2026

 To: Autorità Garante della Concorrenza e del Mercato (AGCM)

Piazza Giuseppe Verdi, 6/A

00198 Roma RM

— and —

 Garante per la protezione dei dati personali

Piazza Venezia, 11

00187 Roma RM

Re: Unfair Commercial Practice and GDPR Transparency — Anthropic PBC (claude.ai) — Opaque Automatic Billing and contract breach by blocking access within cancelled subscription period.

To the competent authorities,

I write jointly to the AGCM and the Garante, as the matters raised engage both unfair commercial practices under the Codice del Consumo and data processing transparency obligations under the GDPR. The underlying conduct is the same; the legal frameworks are complementary.

The Practice

Anthropic PBC, the US-based provider of the AI assistant service Claude (claude.ai), operates a subscription model with supplementary usage-based top-up charges. These charges are triggered automatically against the subscriber's payment card at thresholds that are:

• undefined in advance in any contractual document;
• expressed in internal computational units ("tokens") with no plain-language equivalent;
• applied at Anthropic's sole discretion, without prior notice to the subscriber. 

In practice, the subscriber authorises recurring card charges without knowing when those charges will be triggered, what unit of consumption triggers them, or what the upper limit of liability in any given period might be.

For the AGCM: Unfair Commercial Practice

Article 20 of the Codice del Consumo prohibits unfair commercial practices contrary to professional diligence that materially distort the economic behaviour of the average consumer. Article 22 (misleading omissions) applies directly: Anthropic omits material information — the precise conditions under which automatic charges are levied — that the average consumer requires to take an informed transactional decision.

EU Directive 93/13/EEC on unfair contract terms further provides that terms creating a significant imbalance to the detriment of the consumer are potentially voidable. A clause permitting a supplier to charge at an undefined threshold for an undefined unit of resource is a strong candidate.

The plain language requirement of EU consumer law is also engaged. The operational reality — "we will charge your card when we determine an internal threshold has been crossed" — fails that test categorically.

For the Garante: GDPR Transparency

Processing a payment card transaction requires processing personal and financial data. Under GDPR Article 6, such processing requires a lawful basis. Where Anthropic relies on contractual necessity (Article 6(1)(b)), the contract must be sufficiently transparent for the data subject to understand what processing they are consenting to.

GDPR Article 5(1)(a) requires processing to be lawful, fair, and transparent. Article 13 requires disclosure in concise, intelligible, plain language. A payment trigger that is undefined — exercised at Anthropic's discretion against an undisclosed computational threshold — cannot satisfy these requirements. The data subject has no means of anticipating when their financial data will be processed for a charge, or in what amount.

Pattern of Conduct

The billing opacity is not an isolated contractual anomaly. It reflects a consistent pattern of consumer-hostile service design:

No Service Level Agreement. Anthropic provides no availability guarantee. The industry gold standard for enterprise services is 99.999% uptime ("five nines") — approximately five minutes of permitted downtime per year before penalties apply. Anthropic offers no such commitment and no compensation mechanism for outages, of which there have been multiple. The subscriber is directed to exercise patience.

No Key Performance Indicators. There are no KPIs governing response quality, session continuity, or resource allocation per query. The subscriber bears the full cost of service degradation, including the time and effort required to recover a session when the system loses context — a not infrequent occurrence.

A subscription that implies a relationship but delivers a stranger. Despite the subscription model implying continuity, each session begins stateless — no memory of prior exchanges unless the subscriber manually reconstructs context. The subscriber pays for the accumulated value of an ongoing professional relationship. They receive, on each occasion, a well-briefed stranger. This structural mismatch between what is implied and what is delivered is relevant to the assessment of misleading commercial practices.

My Specific Case

I am an EU-resident subscriber based in Italy. Following cancellation of my subscription, Anthropic charged my card for a top-up. The charge was applied after cancellation was confirmed. No refund has been issued despite contact with Anthropic's support team.

I note that I have previously filed a GDPR Subject Access Request under Articles 15 and 20 against Anthropic in a separate matter, identifying this authority as the competent supervisory authority under Article 77 GDPR.

Request

I request that both authorities investigate Anthropic's automatic top-up charging mechanism and associated disclosure practices. Specifically:

— AGCM: whether the practice constitutes an unfair commercial practice or misleading omission under the Codice del Consumo, and whether corrective measures or sanctions are appropriate;

— Garante: whether Anthropic's payment processing practices comply with GDPR transparency and lawful basis requirements, and whether its billing disclosures meet the plain language standard required of data controllers operating in the EU.

I am happy to provide further documentation on request.

Note: I have chosen this moment to communicate with you because Anthropic have blocked my access to Claude following my written complaint to them yesterday, 7 March 2026. I had fairly normal service until about to 05:30 hours today, 7 March 2026. While Claude down detector shows significant delays, it was present throughout my session until 05:30 this morning. 

My cancelled subscription runs to 20 March 2026. Anthropic are in breach for blocking my access before the contract expiry date.

Yours faithfully,

0808