AI TRANSPARENCY

1. OUR USE OF ARTIFICIAL INTELLIGENCE

TooLong.XYZ uses artificial intelligence (AI) as the core technology behind our video summarization service. We believe in being transparent about how AI is used so you can make informed decisions about using our service.

This page explains what AI does in our product, how your data is handled during AI processing, and the limitations you should be aware of.

2. HOW AI WORKS IN TOOLONG.XYZ

When you submit a YouTube video URL, here is what happens:

1. Transcript Extraction: We retrieve the video's transcript (captions/subtitles) from YouTube using multiple fallback methods to ensure reliability.
2. AI Processing: The transcript is sent to a third-party large language model (LLM) with a carefully crafted prompt that instructs it to generate a structured, accurate summary.
3. Summary Generation: The AI produces a summary that includes key points, main arguments, and important details from the video.
4. Caching: The generated summary is stored so subsequent requests for the same video are served instantly without additional AI processing.

No human reviews individual summaries before they are shown to you. The output is generated entirely by AI.

3. AI PROVIDERS AND MODELS

We use a multi-provider architecture with automatic failover to ensure high availability:

• Primary: Google Gemini (via OpenRouter) — used for the majority of summary generation
• Fallback: OpenAI GPT-4o-mini — activated if the primary provider is unavailable
• Last Resort: Anthropic Claude Haiku — activated if both primary and fallback providers are unavailable

All providers are bound by data processing agreements that prohibit them from using your data for model training or any purpose other than generating the requested summary.

4. DATA HANDLING IN AI PROCESSING

We take your data privacy seriously during AI processing:

• Input Data: Only the video transcript and metadata (title, description) are sent to AI providers. Your personal information (name, email, account details) is never included in AI requests.
• No Training: Your data is not used to train, fine-tune, or improve any AI models. This is contractually enforced with all providers.
• Temporary Processing: AI providers process your data in real-time and do not retain it after generating the summary.
• Encrypted Transit: All data sent to AI providers is encrypted using TLS 1.2 or higher.

5. AI LIMITATIONS AND ACCURACY

AI-generated summaries are helpful tools but have important limitations:

• Accuracy: Summaries may contain factual errors, misinterpretations, or omissions. They should not be treated as a substitute for watching the original video, especially for critical decisions.
• Bias: AI models may reflect biases present in their training data. Summaries may inadvertently emphasise certain perspectives over others.
• Context: AI may miss nuance, sarcasm, cultural references, or visual elements that are not captured in the transcript.
• Technical Content: Highly specialised or technical content may be oversimplified or inaccurately summarised.
• Languages: Summary quality varies by language. English content produces the most reliable results.

We recommend always verifying important information from summaries against the original video or other authoritative sources.

6. NO AUTOMATED DECISION-MAKING

TooLong.XYZ does not use AI to make decisions that have legal or similarly significant effects on you. Our AI is used solely to generate video summaries as a convenience tool. No profiling, scoring, or automated decision-making is performed on your personal data.

7. OUR COMMITMENT TO RESPONSIBLE AI

We are committed to using AI responsibly:

• Transparency: We clearly disclose when content is AI-generated and explain how our AI works
• Privacy First: We minimise the data sent to AI providers and never share personal information
• Continuous Improvement: We regularly evaluate summary quality and update our prompts and provider selection to improve accuracy
• User Control: You choose which videos to summarise and can delete your summaries at any time
• Compliance: We monitor regulatory developments including the EU AI Act and Singapore's AI governance framework to ensure ongoing compliance

8. CONTACT US

If you have questions about our use of AI, want to report an inaccurate summary, or have concerns about AI-generated content, please contact us at:

Email: hello@toolong.xyz