Child safeguarding

Public commitment · draft pending legal review · last updated July 2026

Our commitment

We work with children in institutional care. That makes child protection a precondition, not an appendix: no workshop runs, no mentor speaks to a child, no photo is taken until the protections on this page are in place. Every child’s safety comes before every program goal, every partnership, and every deadline — without exception.

Our policy follows the Keeping Children Safe international standards — policy, people, procedures, and accountability — and complies fully with India’s framework: the Juvenile Justice (Care and Protection of Children) Act 2015, the POCSO Act 2012, and the Digital Personal Data Protection Act 2023.

Who this covers

Everyone acting for CanvasCode, in person or online: staff, trainers, volunteer mentors, board members, contractors, partners, and the founder. Nobody is senior enough to be exempt.

Safe people

• Background verification before any contact with children — police verification, reference checks, and a signed self-declaration, refreshed periodically. This applies to everyone, including the founder and the board.
• Safeguarding training first — no adult interacts with a child before completing training on this policy, and every adult signs the code of conduct annually.
• The two-adult rule — no unsupervised one-to-one contact between an adult and a child, physical or digital. Mentoring happens on supervised, logged channels; sessions happen in open, visible spaces.
• No private contact — adults never contact children through personal phones, personal social media, or private messaging, and never request a child’s personal contact details.

Safe programs inside partner organizations

• We work only inside child care institutions registered under Section 41 of the JJ Act, and we verify registration before partnership.
• Programs run with the formal permission of the institution and, where required, in coordination with the district Child Welfare Committee — the statutory authority over the care of children in institutions.
• Every program activity is risk-assessed before it runs, and each child’s participation respects their individual care plan and requires documented guardian consent and the child’s own assent, always.

A safe digital platform, and how we use AI

• No social features. Our program application has no child-to-child messaging, no comments, no public profiles, no leaderboards. A child sees their own work.
• AI is supervised, always. AI may draft feedback on a child’s work, but a trained adult reviews, edits, and approves it before the child ever sees it. The AI never speaks to a child directly and never decides anything about a child.
• Data minimization. Student accounts use a first name and initial only — no surname, no date of birth beyond an age band, no photograph. Children’s data is never used for tracking, behavioural monitoring, advertising, or training AI models (DPDP Act, Section 9(3) — and our own stricter bar).
• Supervised tools. Children use AI tools on supervised accounts with filtered access, and every track teaches digital safety inside the missions.

Photos, stories, and children’s privacy

No identifiable child appears in our public materials — no names, faces, or identifying details. For children in institutional care this is also the law: Section 74 of the JJ Act prohibits publishing anything that could identify a child in a child care institution. We show the work, never the face, and we tell stories that grant children the dignity of their fight — never pity.

Mandatory reporting

If anyone in our organization learns that a child may have been sexually abused, Indian law — the POCSO Act, Sections 19 to 21 — requires reporting it to the police or the Special Juvenile Police Unit. We will always do so, and we will never investigate internally instead of reporting. Failing to report is itself an offence, and we treat the duty as seriously as the law does.

How children can speak up

Children in our programs are told, at induction and in their own language, how to raise a concern: to our safeguarding lead through any trusted adult or through the program app, to their institution’s management, to the Child Welfare Committee — or to ChildLine 1098, India’s 24-hour helpline for children, at any time. No child will ever face consequences for speaking up.

How adults can raise a concern

Write to hello@canvascode.org marked “Safeguarding” — it reaches the safeguarding lead directly. Reports are handled confidentially, whistleblowers are protected, and failing to pass on a concern is treated as a disciplinary matter. A named safeguarding lead and deputy will be published here when the founding team is complete.

Accountability

• This policy is reviewed annually and after every incident or near-miss.
• A board-level safeguarding owner reports on implementation at every board meeting.
• We self-assess against the Keeping Children Safe standards and commit to periodic external review, with findings reported in our annual report.
• Honesty with partners and authorities always takes precedence over reputation management.

The full signed policy document will be published on the transparency page alongside our registrations. Questions? Ask us directly.