Transparency
AI Nutrition Labels — Built on Trust
Whether you're a school, a trust or a business, you deserve to know what's under the hood. Our AI Nutrition Labels give you a clear, honest breakdown of how the platform works — from data handling and hosting to safeguarding and risk.
LLM providers: OpenAI · Google · Anthropic · Perplexity. Trained on your data: No — your data is never used to train public models. Data sent to model: Only the text of the user's query and any documents deliberately uploaded. External web access: Not by default — can be enabled per admin configuration.
Framework: UK GDPR · DfE aligned · KCSIE aligned. Hosting region: UK — AWS eu-west-2 (London). PII handling: User identifiers removed during day-to-day processing. Data sold/shared: Never. Analytics sharing: Anonymised usage insights only, if opted in. Data retention: Retained for duration of service; permanently deleted on account closure.
Automated content flagging engine active on all conversations. Alerts sent automatically to your organisation's designated trust admin contact when triggered. Statutory duty to escalate where there is risk of danger to self or others, in line with KCSIE obligations. Statutory guidance: KCSIE · SEND Code of Practice · Ofsted SIH · Teacher Standards · Pupil Premium.
All outputs are advisory only. Staff must review and apply professional judgement before acting or sharing. Full interactive AI Nutrition Labels (Education & Organisations editions) available at our Transparency page.