| Deployer | Easylab AI |
|---|---|
| Service | LinkedInScope |
| AI Act Classification | Non-high-risk (Article 6(3)) |
| Transparency Obligations | Article 50 |
| Last Updated | March 17, 2026 |
| Provider | OpenAI |
|---|---|
| Model | GPT-4o-mini |
| Type | Large Language Model (text generation) |
| Purpose | Analyze LinkedIn profiles and generate structured feedback reports |
| Input Data | Publicly available LinkedIn profile data (name, headline, about section, experience, posts) |
| Output | Structured JSON analysis with identity analysis, consistency check, and recommendations |
| Data Retention | Analysis results stored in Firestore; profile data not retained beyond the analysis session |
| Training | No user data is used for model training (OpenAI API usage) |
| Limitations | The analysis represents AI-generated opinions and may contain inaccuracies or biases inherent to the model. Results should not be taken as professional career advice. |
| Risk Mitigation | Outputs are clearly labeled as AI-generated. The system character ("Monday") signals through its tone that the analysis is not a factual assessment. |
| Provider | Google (Gemini) |
|---|---|
| Models | Gemini 2.5 Pro (photo analysis), Gemini image generation (image creation) |
| Type | Multimodal AI (image analysis + generation) |
| Purpose | Generate stylized profile photos from user-uploaded selfies |
| Input Data | User-uploaded photograph (selfie), style preferences, social network selection |
| Output | AI-generated stylized profile photo (2K resolution JPEG) |
| Data Retention | Source images stored temporarily for processing; generated images stored for user download |
| Watermarking | AI-generated images contain metadata watermarks applied by Google Gemini, in compliance with Article 50(2) |
| Training | No user images are used for model training |
| Limitations | Generated images are artistic interpretations and may not perfectly represent the subject. Style application quality varies. |
| Risk Mitigation | Clear labeling of AI-generated content, metadata watermarking, user consent required before generation |
Users are informed about AI usage through the AI Governance page and this documentation.
Clear notification that users interact with AI systems.
Disclosure that content is AI-generated. AI-generated images contain metadata watermarks.
Processing in accordance with Regulation (EU) 2016/679.
For questions about our AI systems or to exercise your rights:
support@linkedinscope.com