Safety & trust

Safe for children is the product, not a feature

Turing Education is built to the Department for Education's Generative AI product safety standards (updated January 2026). Here is how, in plain English.

Filtering

Every message is checked by fixed rules and an AI second opinion before the AI answers — including misspellings and patterns across a session. Stricter in Junior mode.

Monitoring & reporting

Every conversation is saved. Teachers see a plain-English class summary. Pupils are told clearly that adults can see their chats.

Safeguarding

Serious concerns go to your DSL and deputy, not just the class teacher. Pupils can't join until the DSL's details are added.

Thinking, not copying

Hints first. Full answers only after a genuine attempt. Teachers can switch final answers off entirely.

Healthy use

A daily time limit set by the teacher. When it's reached, Ask stops until tomorrow. No streaks or tricks to keep children online.

No AI 'friend'

No name, no face, no 'I feel'. The AI is a tool and always points children to real people.

Privacy & data

Pupil work is never used to train AI. Pupil names are hidden in safeguarding lists until a DSL opens a case.

Accountability

Every concern has a timeline, notes and a record of each time it was viewed.

Serious concerns appear in the DSL’s Turing Education inbox straight away and are emailed to the DSL and deputy. Text-message alerts are coming soon.