AI Policy

Dear Future Me uses artificial intelligence (AI) to enhance your letter-writing experience. We are committed to transparency about how AI is used in our app, what data is processed, and how your privacy is protected.

This AI Policy supplements our Privacy Policy and Terms of Service. It is designed to comply with the EU AI Act (Regulation (EU) 2024/1689), GDPR, and applicable AI transparency requirements.

We believe you have the right to understand when and how AI interacts with your personal content. This document provides that transparency.

Dear Future Me incorporates three AI-powered features:

All AI features are powered by Anthropic's Claude API.

What It Does

When you open a sealed letter, Dear Future Me can generate an AI-powered reflection that provides thoughtful commentary on the themes, emotions, and personal growth reflected in your letter. This feature is designed to enrich the experience of revisiting your past thoughts.

How It Works

1. You open a sealed letter that has reached its unlock date.
2. The app presents the option to generate an AI reflection.
3. If you choose to proceed, the following data is sent to Anthropic's Claude API:
   • The text content of your letter
   • The mood you selected when writing the letter
   • The date you wrote the letter
   • The current date
4. Claude API processes this data and returns a personalized reflection.
5. The reflection is stored in your letter record on our servers.
6. Anthropic does NOT retain any of the data after processing.

What AI Reflection Is NOT

• It is NOT therapy or psychological counseling.
• It is NOT professional advice of any kind.
• It is NOT a diagnostic tool for mental health conditions.
• It should NOT be used as a substitute for professional help.

What It Does

When you seal a letter, Dear Future Me generates a unique piece of visual art that serves as your letter's seal. Each seal is one-of-a-kind, reflecting the mood and essence of your letter.

How It Works

1. You complete writing your letter and choose to seal it.
2. The following data is sent to Anthropic's Claude API:
   • The mood you selected for your letter
   • The general theme/category of the letter (e.g., “gratitude”, “hope”, “reflection”)
   • The seal duration selected
3. Claude API generates a visual description or prompt for the seal art.
4. The generated seal art is stored in our seal-art storage bucket (publicly accessible as decorative art, not linked to personal content).
5. Anthropic does NOT retain any of the data after processing.

Important Note

• The actual text content of your letter is NOT sent to the AI for seal art generation. Only the mood and thematic metadata are used.
• Seal art images are decorative and do not contain or encode any of your personal letter content.

What It Does

Dear Future Me provides AI-generated writing prompts to inspire you when you are looking for ideas about what to write. These prompts are general and creative in nature.

How It Works

1. Writing prompts are generated periodically by VietByte using Anthropic's Claude API.
2. Generated prompts are cached in our database (writing_prompts table) and served to all users.
3. Prompt generation uses aggregated, anonymized patterns such as:
   • Popular themes across the user base
   • Seasonal and temporal relevance
   • General writing categories
4. Your individual letter content is NEVER used to generate writing prompts.
5. Writing prompts are the same for all users in a given period — they are not personalized to your specific content.

Important Note

• Writing prompts are pre-generated and cached. No real-time AI processing of your data occurs for this feature.
• No personal data is sent to the AI for prompt generation.

The following table summarizes exactly what data is sent to Anthropic's Claude API for each feature:

Technical Details

• All data is transmitted via encrypted HTTPS connections (TLS 1.2+).
• Data is sent from our Supabase Edge Functions (server-side) to Anthropic's API — your device does not communicate directly with Anthropic.
• API calls use Anthropic's zero-retention API tier, meaning submitted data is not stored by Anthropic after the response is generated.
• Each API call is stateless — no conversation history or context is maintained between calls.

We want to be explicitly clear about what Anthropic's AI does NOT do with your data:

Anthropic Does NOT:

• Store your letter content, mood data, or any personal information after processing the API request.
• Use your data to train, fine-tune, or improve AI models. Your data is excluded from training under our API agreement.
• Share your data with any third parties.
• Build user profiles or behavioral models based on your data.
• Retain logs of your data beyond the immediate API request processing.
• Access your images, voice recordings, or handwriting data (these are never sent to the AI).

VietByte Does NOT:

• Send your data to any AI provider other than Anthropic.
• Use your data for AI training or model improvement.
• Allow AI to make automated decisions that significantly affect you.
• Use AI for profiling, scoring, or categorizing users.

Contractual Protections

Our agreement with Anthropic includes:

• A zero-retention data processing commitment.
• Prohibition on using API data for model training.
• Data processing terms compliant with GDPR Article 28.
• Standard Contractual Clauses (SCCs) for international data transfers.

You have the right to opt out of AI features at any time, without any impact on the core functionality of Dear Future Me.

How to Opt Out

Global opt-out: Settings > Privacy > AI Features > Toggle off

What Happens When You Opt Out

What Does NOT Change

• You retain full access to letter writing, sealing, opening, and all core features.
• Your existing AI-generated reflections and seal art are NOT deleted (unless you request deletion).
• You can re-enable AI features at any time.
• Your premium subscription benefits (if any) are not affected.

AI-generated content in Dear Future Me is provided for personal enrichment and creative inspiration. Please be aware of the following:

AI Reflections

• AI reflections are generated by a language model and may not always accurately capture the nuance or intent of your letter.
• Reflections may occasionally contain generic, irrelevant, or unexpected observations.
• AI does not have context about your life beyond what is contained in the individual letter being processed.
• AI reflections are NOT professional psychological, medical, or therapeutic assessments.
• If you are experiencing mental health difficulties, please contact a qualified professional. AI reflections are not a substitute for professional care.

AI Seal Art

• Generated seal art is creative and abstract. Its interpretation is subjective.
• Seal art quality and style may vary.

AI Writing Prompts

• Writing prompts are general suggestions and may not always be relevant to your current situation.
• Prompts are generated by AI and may occasionally contain unexpected or imperfect suggestions.

General

• VietByte does not guarantee the accuracy, appropriateness, or quality of any AI-generated content.
• AI outputs should be treated as creative suggestions, not authoritative statements.
• We continuously work to improve AI output quality, but imperfections are inherent to current AI technology.

Dear Future Me's AI features are classified under the EU AI Act (Regulation (EU) 2024/1689) as follows:

Risk Classification

Our AI features are assessed as minimal risk / limited risk systems under the EU AI Act:

• AI Reflection: Limited risk — transparency obligations apply. Users are informed that content is AI-generated and can opt out.
• AI Seal Art: Minimal risk — generates decorative visual content with no significant impact on users.
• AI Writing Prompts: Minimal risk — provides general creative suggestions with no personalization based on individual data.

Transparency Obligations (Article 50)

In compliance with EU AI Act transparency requirements:

• Users are clearly informed that reflections, seal art, and writing prompts are generated by AI.
• AI-generated content is labeled as such within the app interface.
• This AI Policy provides comprehensive disclosure of AI functionality, data processing, and user rights.
• Users can distinguish between their own content and AI-generated content at all times.

Our AI Does NOT:

• Engage in prohibited AI practices (Article 5) such as subliminal manipulation, exploitation of vulnerabilities, social scoring, or biometric categorization.
• Constitute a high-risk AI system (Annex III) as it does not perform risk assessment, profiling, or automated decision-making in sensitive areas.
• Generate deepfakes or synthetic media that could be mistaken for real content.

We may update this AI Policy as our AI features evolve or as regulations change. When we make material changes:

• We will update the “Last Updated” date at the top of this policy.
• We will notify you via in-app notification at least 14 days before changes take effect.
• For changes that affect how your data is processed by AI, we will request renewed consent where required by law.

If you have questions about our AI practices:

VietByte Company Limited
Email: privacy@vietbyte.tech
Data Protection Officer: dpo@vietbyte.tech

For AI-specific concerns, please include “AI Policy” in your email subject line.