How We Use AI
SoScripted uses AI at two stages of transcript processing:
- Transcription — Our proprietary transcription pipeline processes video from 7 social platforms, combining platform-native caption extraction with our own speech-to-text processing for maximum accuracy and coverage.
- Content analysis — After transcription, our AI models extract topics, entities (people, companies, products, tools), and key takeaways from each transcript. We also generate vector embeddings for semantic search across your library.
Data Handling
- Transcript text is processed by our AI pipeline for tagging and embedding generation. Your data is never used for model training by any party.
- Embeddings are stored in our encrypted database and are private to your account.
- We do not sell, share, or use your transcript data for any purpose other than providing the service to you.
Transparency
All AI-generated metadata is clearly labeled in the product:
- Topics, entities, and key takeaways are displayed in an "Insights" section, separate from the original transcript text.
- Semantic search results include a similarity score so you can judge relevance.
- The original transcript text is never modified by AI processing.
Responsible Use
Our API terms of service prohibit using SoScripted for:
- Surveillance or monitoring individuals without their knowledge
- Harassment, doxxing, or targeting individuals
- Generating misleading content attributed to video creators
- Circumventing platform-specific content restrictions or copyright protections
We reserve the right to suspend accounts that violate these terms.