AI Policy

Effective date: 1 June 2026 Version: 2026-06-01

This AI Policy describes how the PMGclub platform processes AI feature inputs and outputs, which AI providers it uses, and what your rights and obligations are when using them.

This Policy supplements the Terms of Service and Privacy Policy.


1. What AI Features the Platform Offers

The PMGclub platform uses artificial intelligence to provide the following features:

1.1 Text

  • Text generation — post suggestions, summarisation, wording improvements
  • Translations — automatic translation of posts and presentations between Czech, English, German, and other languages
  • Hashtag suggestions — automatic recommendation of relevant tags
  • AI search — semantic search across platform content

1.2 Images

  • Image generation — AI images for post covers, presentations, profiles
  • AI Vision (Anthropic Claude) — recognising the content of uploaded images for the "AI Paste" feature

1.3 Audio

  • Voice generation (TTS) — narrating text in various voices and languages for presentations
  • Background music — generating instrumental music for presentations

1.4 Complex tools

  • Smart Presentation Generator — AI generates structure and content of a presentation based on a brief
  • Pekari (AI notebook) — full content generation from uploaded sources (URL, DOCX)
  • QLive AI — automatic generation of quiz questions and summaries from a presentation

2. AI Providers

The platform uses services from the following providers:

Provider Services used Location
OpenAI, L.L.C. GPT-4o-mini (text), TTS (voice), embeddings, DALL-E (images) USA
Anthropic, PBC Claude (text, AI Vision) USA
ElevenLabs Inc. Advanced TTS for presentation narration USA

All providers are either certified under the EU-US Data Privacy Framework or process data on the basis of Standard Contractual Clauses (SCCs) approved by the European Commission.

3. What Is Sent to AI Providers

3.1 Inputs

When you use an AI feature, the following is sent to the provider:

  • your prompt or brief (the text you write)
  • the content you want to process (an existing post to improve, text to translate, a document to summarise)
  • uploaded files (if the feature requires an image input — e.g. AI Vision)
  • technical request metadata (operation type, model, language)

3.2 What Is NOT Sent

  • your name, email, or user ID
  • your billing data
  • history of other interactions with the platform
  • credentials

The operator strives to minimise personal data transmitted to AI providers. If you yourself include personal data in the prompt (e.g. names, contacts), you, as the author of the input, are responsible.

4. AI Model Training

4.1 OpenAI and Anthropic

The operator uses the API tier of OpenAI and Anthropic services, where, under the providers' terms, inputs and outputs are not used to train models.

4.2 ElevenLabs

ElevenLabs' terms do not allow the same unambiguous guarantee. Voice inputs (text you narrate) may be retained for the time necessary to operate the service under the ElevenLabs Privacy Policy.

5. Your Obligations When Using AI Features

5.1 Do not enter sensitive data

We strongly advise against entering into AI features:

  • sensitive personal data of third parties (name + health status, financial situation, sexual orientation, political opinions),
  • data about minors,
  • credentials, passwords, API keys,
  • bank data and payment card data,
  • third-party trade secrets,
  • classified information under special laws.

Entering such data may violate GDPR, banking secrecy, or professional confidentiality obligations. The operator is not liable for such violations.

5.2 Respect AI providers' terms

When using AI features, you must respect the Acceptable Use Policy of the relevant provider:

Generally prohibited include: generation of illegal content, deepfakes of persons without consent, materials exploiting minors, election manipulation, public health disinformation, weapons manufacturing instructions, malware.

Violating an AI provider's terms may result in suspension of AI feature access across the entire platform — the penalty is not individual to a single user but affects all platform users.

5.3 Label AI content where it could mislead

If AI generates content that could be mistaken for an authentic record of a real event (quotes of persons, photos, voice recordings), label it as generated by artificial intelligence.

Details in the Community Guidelines, Article 4.

6. AI Outputs and Your Rights

6.1 No accuracy warranty

AI outputs are provided without warranty of accuracy, completeness, timeliness, or fitness for any particular purpose. Models may:

  • generate factually incorrect information ("hallucinations"),
  • omit important context,
  • exhibit bias from training data,
  • generate content that looks credible at first glance but is not.

Always verify AI outputs before publishing or using them in an important context.

6.2 No professional advice

AI outputs do not constitute professional advice — financial, legal, medical, tax, or otherwise. For important decisions, consult a qualified professional.

6.3 Rights to AI outputs

Rights to use AI outputs are governed by the terms of the relevant AI provider. In general:

  • OpenAI — the user acquires rights to the extent granted by OpenAI (Terms of Use)
  • Anthropic — outputs belong to the user, subject to the AUP (Commercial Terms)
  • ElevenLabs — voice content may be used commercially under Terms

The platform operator does not claim ownership rights to AI outputs you generate via platform features, except for a non-exclusive licence to display them on the platform (see Terms of Service, Article 11).

6.4 Copyright notice

AI-generated content does not enjoy copyright protection in the Czech Republic or most EU jurisdictions unless it contains significant human creative input. This means purely AI-generated text or images can be used by anyone, and you cannot protect them under copyright.

7. Credit Consumption and Transparent Pricing

Each AI operation has a transparent price in credits, displayed before the operation is performed. The operator may:

  • Publish the current pricing of operations at pmgclub.com/credits/pricing
  • Change the price of operations in response to changes in AI providers' prices — we will inform you of material changes 30 days in advance
  • Bill overage of the monthly allocation via a separate invoice if you have enabled this option (see Terms of Service, Article 7.5)

8. Security Measures

The operator applies the following measures when using AI:

  • Rate limiting — limiting the number of requests per time unit (abuse prevention, protection against excessive credit consumption)
  • Content filtering — filtering of prompts that could violate the AI provider's AUP
  • Logging — AI usage records (who, when, which operation, how many credits) are stored for 12 months for billing, debugging, and security audit purposes
  • Input sanitisation — removal of prompt injection attempts from user inputs where possible

9. What to Do if There Is a Problem

9.1 Incorrect or inappropriate AI output

If the AI generates an output that is:

  • factually incorrect — do not use the output; in the operating interface you can give a "thumbs down", which helps improve the system
  • offensive or inappropriate — contact us at info@pmgclub.com with the operation ID so we can investigate
  • violates your personal rights or rights of third parties — contact us as soon as possible and we will remove the output from the platform

9.2 Technical problem

AI features may be temporarily unavailable due to: - AI provider service outage - exceeded rate limits - platform maintenance

In such cases, credits that would have been consumed are refunded (if they have not already been successfully used).

10. Changes to this Policy

The operator will inform users of material changes to this Policy (adding a new AI provider, substantial change in processing scope) at least 14 days in advance via email or in-app notice.

11. Contact

PMG Capital s.r.o. Email: info@pmgclub.com


Effective from: 1 June 2026 PMG Capital s.r.o. · Bohuslava Niederleho 1018, 272 04 Kladno, Czech Republic · Company ID: 06998771