AI Content Policy

Coffeescribe is an AI-powered platform operated by 7sumcreations LLC(a Wyoming limited liability company, registered address: 1309 Coffeen Ave, Suite 1200, Sheridan, Wyoming 82801, USA) — artificial intelligence is central to everything we do. This policy explains how AI generation works, what it means for your content and data, and what responsibilities you take on as a creator. Please read it carefully before creating and publishing books.

1. How AI Generation Works

When you create a book on Coffeescribe, your prompts, topic, and settings are sent to third-party AI language models to generate the content. We do not train or host our own AI models. Instead, we route requests through OpenRouter, which connects to a range of AI providers including Anthropic, Google, Meta, and others.

The AI model available to you depends on your subscription tier:

• Free tier: Budget models — lower cost, capable, but prompts may be used for training by model providers
• Pro tier: Standard models — better quality, prompts not used for training
• Creator tier: Premium and advanced models — highest quality, prompts not used for training

We may change the specific models available at any tier at any time to improve quality, manage costs, or respond to changes in model availability. The current model selection is visible in the book creation form.

2. Free Tier Data Sharing

Free tier AI models are provided at no cost by third-party providers who fund this through data sharing arrangements. By using the free tier, you acknowledge and agree that:

• Your book prompts and generated content may be used by those model providers for AI training
• Your inputs and outputs may be published by those providers to public datasets (anonymously)
• Free model providers may retain and log your prompts and responses
• This provider-side sharing cannot be opted out of while using the free tier

What is never shared: Your account information, email address, and personal details are never shared with AI model providers, regardless of tier. Only your book creation prompts and AI-generated responses may be used as described above.

Coffeescribe itself never trains on your content. Everything described in this section happens at the third-party model provider, not at Coffeescribe. We do not retain prompt text or AI responses on our servers for training purposes, and we do not use your prompts, books, research notes, or uploaded documents to train our own models — on any tier. See the Privacy Policy (AI Training) for full details.

3. Accuracy & Hallucination Risks

AI language models generate plausible-sounding text — but plausible is not the same as accurate. Common issues include:

• Hallucinations: AI may invent facts, statistics, citations, quotes, or references that do not exist
• Outdated information: AI models have a knowledge cutoff and may not reflect recent events
• Biased perspectives: Outputs may reflect biases present in training data
• Inconsistencies: Longer books may contain contradictions across chapters
• Plausible-sounding errors: Incorrect information may be presented confidently

AI-generated content must not be relied upon as professional, medical, legal, financial, psychological, or any other form of expert advice. Always consult qualified professionals for decisions that matter.

You are responsible for reviewing, fact-checking, and editing all content before publishing it or sharing it with others.

4. Content Ownership

You retain ownership of the books you create on Coffeescribe, including your prompts, edits, and any original contributions. However, there are important limitations to be aware of:

• No exclusivity: Similar prompts can produce similar outputs for different users. Coffeescribe does not and cannot guarantee that your generated content is unique
• Copyright uncertainty: The copyright status of AI-generated content is an evolving legal area that varies by jurisdiction. You are responsible for understanding the applicable laws in your region before publishing or commercialising content
• Training influence: AI models are trained on existing content, which may influence outputs. We cannot guarantee that generated content does not resemble existing works
• Your responsibility: You are responsible for ensuring your use of generated content complies with all applicable laws, including copyright, defamation, and consumer protection laws

When you make a book public, you grant Coffeescribe a non-exclusive licence to display and distribute that content within the platform. See our Terms of Service for full details.

5. Content Moderation

We do not pre-screen AI-generated content before it is shown to you or published. AI models may occasionally produce content that is inaccurate, inappropriate, or unexpected. You are responsible for reviewing all generated content before making it public.

Our prohibited content policy applies to all books published on the platform, whether AI-generated or user-edited. This includes content that is illegal, sexually explicit, hateful, defamatory, or otherwise harmful.

Community reporting:Any user can report a public book using the "Report Content" button. The button opens a modal with a pre-filled email template containing the Book ID, title, and URL — you copy the template (or click "Open Email Client") and send it to support@coffeescribe.ai with subject "REPORT CONTENT" (in all caps), adding the reason and any context. We review all reports within 24–48 hours and will remove content that violates our terms.

Brew Preview content (anonymous generation):The homepage Brew Preview widget generates content for visitors with no account. This content is shown only to the visitor in their browser session and is never published to the library. If a visitor joins the waitlist and later creates an account, the brewed content is converted into a private draft scribe visible only to them — it does not become a public book automatically. The standard “Report Content” flow in the library covers published public scribes. Concerns about brew preview content should be directed to support@coffeescribe.ai.

Content appeals: If your content has been removed and you believe the decision was made in error, email support@coffeescribe.ai with the subject line "CONTENT APPEAL", including the book title and reason for appeal. We will review your appeal and respond as soon as possible.

6. Model Transparency

We believe you should know what AI is generating your content. Here is how models are assigned per tier:

The specific models available in each category may change at any time. The current selection is always visible in the book creation form. We aim to give paid users access to the best available models as they are released.

AI models are provided by third-party companies including Anthropic, Google, Meta, Mistral, and others, accessed via OpenRouter. Each provider has their own terms and privacy policies governing model behaviour and data handling.