Safeguarding policy for Split Happens

1. Purpose and scope

The purpose of this policy is to safeguard and promote the welfare, rights, and safety of users of Split Happens, particularly those aged 13–30 who use our website and AI support services. It applies to all content, interactions and technical systems within the platform including the AI chatbot, videos of personal stories and links to trusted resources and services.

2. Our commitment

We are committed to:

• Promoting a safe, supportive environment for children and young adults
• Identifying and appropriately responding to safeguarding risks through clear signposting
• Ensuring our AI and human content follow responsible, trauma-informed and youth-sensitive practices
• Providing clear pathways to professional support when needed

3. Safeguarding objectives

These are to:

• Prevent harm or abuse to users interacting with our services
• Promote mental and emotional wellbeing
• Recognise and respond appropriately to concerns around suicidal thoughts, abuse, exploitation, or neglect
• Educate users about healthy coping strategies and where to find help
• Be in line with UK GDPR and the ICO's Children's Code (Age Appropriate Design Code)
• Be guided by the United Nations Convention on the Rights of the Child (UNCRC), particularly Article 12 where children have the right to express their views freely

4. Key safeguarding risks

• Exposure to triggering or distressing content
• Misuse of the AI chatbot for harmful or misleading advice
• Breaches of data privacy or accidental disclosure
• Underage users accessing the platform
• Emotional overreliance on the platform without live human support
• Bad advice (e.g. cutting off relationships unnecessarily, or remaining in unsafe ones)
• People vulnerable to suicide or self-harm not being signposted to appropriate help

5. Safeguarding measures

a. Access and age restrictions

• The website and AI chatbot are designed for users aged 13 or over.
• Users who indicate they are under 13 are not permitted to use the AI chatbot and are directed to parent-led resources instead.
• Parental guidance is recommended for those under 16.
• The website includes clear wording to remind users that they are communicating with an AI and that the AI has limitations.

b. AI chatbot

• The AI chatbot does not store personal or identifying information on Split Happens servers.
• User messages are processed securely via our AI provider in order to generate responses but are not retained by Split Happens as conversation content.
• Where technical identifiers are used (such as anonymised conversation or session IDs), these cannot be used to identify an individual user.
• The AI does not search the open internet. Its responses are generated only from content provided by Split Happens and reviewed by subject-matter experts, reducing the risk of hallucinations or unsafe advice.
• The AI chatbot does not save or retain full conversation histories.
• The bot is programmed to recognise distress signals or risk-related language such as self-harm, abuse, neglect or suicide and to consistently signpost users to appropriate crisis support services.
• The bot is not a substitute for therapy or emergency services and clearly states this.
• Because Split Happens does not collect identifying or contact information, it cannot make direct safeguarding referrals. Instead, the platform operates a signposting-only safeguarding model, encouraging users to seek appropriate off-platform support.
• In addition to signposting, the bot may ask gentle follow-up questions to help users reflect and decide on next steps for themselves. Additional resources are available elsewhere on the website.

c. Content moderation

• The AI is only grounded in approved materials supplied by Split Happens stakeholders and does not use external or public internet sources.
• Emotional support prompts are used across the website to ensure that when sensitive topics are discussed, users are guided towards further help where appropriate.
• Users can report issues or suggest improvements to content and responses via [email protected].

d. Data protection

• No personal or identifiable information is required to access the AI or resources.
• Users are asked their age during onboarding to ensure age-appropriate responses; this is stored only as an age band and not as an exact age or date of birth.
• Data protection and GDPR compliance are strictly observed and align with the ICO's Children's Code.
• We protect the anonymity of young people who sign up to Split Happens by collecting no personal or identifying details, and instead recommending that they choose an anonymous username.
• We protect the anonymity of young people who share their story by using a fictional first name.
• Any young person who offers their story to be filmed for the site completes a usage and consent form confirming they are happy for the content to be published.

6. Review and updates

This policy will be reviewed annually or in response to major updates in legislation, platform technology, or safeguarding guidance.

7. Contact information

For safeguarding, technical or content queries:

[email protected]

Last review: 09/01/2025

Next review due: 30/06/2026