AI Transparency Notice - BScale AI
Last updated: 2026-05-27
This notice describes our use of generative AI, in line with EU AI Act Art. 50 and global best practices.
Models in use
Text: Anthropic Claude, OpenAI GPT, Google Gemini.
Images: OpenAI (DALL-E / GPT Image), Google Gemini Image.
Video: Runway, Kling, Seedance.
Audio (TTS): a text-to-speech provider.
How outputs are labeled
Every AI-generated asset is stored in your workspace with metadata indicating it was AI-generated and which model produced it.
We recommend adding an 'AI-generated/edited' label to public publications - a requirement that is expanding across jurisdictions.
What we do NOT do with your inputs
We do not sell prompts or assets you create.
We instruct AI providers, where offered, not to use your inputs to train their public models (no-train opt-out).
Sensitive content you upload is processed by providers, so consider what you share.
Known limitations
AI models may generate inaccurate information or hallucinations.
Outputs may reflect biases present in training data.
Responsibility for reviewing outputs and confirming fit-for-purpose rests with you.
Automated decisions
Platform recommendations (e.g., campaign optimization) are decision support only.
Any actual action runs only after you actively approve it in the UI.
Per GDPR Art. 22 - you are not subject to fully automated decisions with legal effects.
Questions
main@bscale.co.il.