ImpactImpact Learning

AI Model Providers

External providers that process your data in Impact Learning

AI Model Providers

Impact Learning uses external AI providers to generate content, images, audio, and to search the web. We have carefully selected providers that meet the security and data protection standards expected in corporate environments. This page gives you a complete overview of where your data is processed, so you can evaluate each provider against your own organisation's requirements.

All External Providers

ProviderWhat It DoesData Sent
AnthropicContent generation (Claude models)Chat messages, source materials, template instructions
OpenAIContent generation (GPT models), audio narration (TTS)Chat messages, source materials, narration scripts
GoogleContent generation (Gemini models), image generation (Gemini Image)Chat messages, source materials, image prompts
InceptionContent generation (Mercury models)Chat messages, source materials, template instructions
ElevenLabsAudio narrationNarration scripts
Exa LabsWeb search for current informationSearch queries

Content Generation

You choose which model to use for content generation. Your selection determines which provider receives your chat messages and source materials.

ProviderModels
AnthropicClaude Sonnet (default), Claude Opus, Claude Haiku
OpenAIGPT-5.4
GoogleGemini 3 Pro, Gemini 3 Flash Lite
InceptionMercury 2

Image Generation

All image generation is handled by Google via the Gemini 3.1 Flash Image model. The AI agent constructs the image prompt based on your course content and sends it to Google for generation.

Audio Narration

Narration scripts are sent to either ElevenLabs or OpenAI depending on the voice provider selected. Only the narration text is sent — not your source materials or chat history.

When the AI agent needs current information, it sends search queries to Exa Labs. Only the search query is sent — not your course content or source materials.

On this page