Its platform, approved informational content, AI behaviour, safeguarding responses, privacy, security, reporting design and product governance.
Safety & governance
Helpful by design.
Safe by design.
Accountable throughout.
Support Never Stops is designed for a sensitive, high-consequence environment. Safeguarding, content controls, privacy, security and human accountability are built around the AI experience—not added after it. The platform is not a clinical system and does not make, manage or influence clinical policy or clinical decisions.
Safety & governance framework
People first
Safe, clear and accountable support
Clear scope
Privacy
Safeguarding
Accountability
Documented controlsRules, hazards and decisions
Human oversightNamed governance ownership
Continuous reviewLearning, monitoring and change
Illustrative framework. The detailed assurance plan is agreed for each deployment and its intended use.
Trust depends on more than a good answer.
People may use Support Never Stops when they are worried, uncertain, isolated or at immediate risk. Our governance approach is therefore built around clear boundaries, proportionate data use, reliable escalation and accountable human decision-making.
A clear division of responsibility
Support Never Stops is not a clinical system.
It provides information, learning, harm-reduction guidance, signposting and service intelligence. It does not provide clinical care and does not create, manage, approve or influence clinical policy or clinical decision-making.
All clinical policy, assessment, diagnosis, prescribing, treatment, professional judgement, care decisions, protocols and local clinical governance.
The boundary is deliberate: Support Never Stops may complement a service, but it does not become part of that service’s clinical decision-making process and does not seek to replace or influence professional judgement.
Six pillars of safe and responsible delivery
Clear non-clinical scope
The platform is designed as an information, learning and support tool. It does not assess, diagnose, prescribe, determine treatment or make decisions on behalf of a service or professional.
Safeguarding and urgent risk
Overdose, immediate danger, abuse, exploitation and other serious concerns trigger clear, pre-defined safety responses and routes to human help.
Privacy and data minimisation
The platform is designed to avoid unnecessary personal information and to separate individual support from anonymised service insight.
Governed content and AI
Approved guidance, explicit behavioural rules, testing and controlled updates shape how the model responds in sensitive situations.
Security and resilience
Access controls, secure hosting, monitoring, incident processes and recovery arrangements protect the service and the information it handles.
Named human accountability
Product, safeguarding and information-governance responsibilities remain with identifiable people and organisations. All clinical governance, policy and decision-making remain entirely with the service.
Safety in the conversation
When risk appears, the response changes.
Support Never Stops is not designed to continue an ordinary information exchange when someone describes an emergency or a serious safeguarding concern. The conversation moves into a controlled safety pathway that prioritises immediate action and human support.
The platform cannot guarantee that every risk will be identified. Safety controls reduce foreseeable risk and are supported by clear user messaging, testing, monitoring and human governance.
1 · Listen
Recognise relevant language and context
The conversation checks for indicators of overdose, immediate danger, self-harm, abuse or exploitation.
2 · Prioritise
Switch to the appropriate safety rule
Potentially unsafe conversational freedom is reduced and the relevant urgent pathway takes priority.
3 · Act
Give clear, practical next steps
The user is directed towards emergency help, overdose response, safeguarding support or another approved route.
4 · Connect
Keep human help visible
The system reinforces that emergency services and qualified professionals—not the AI—must manage urgent care.
Clear boundaries protect users and professionals
Support Never Stops is designed to…
✓
Provide accessible, non-judgemental harm-reduction information.
✓
Ask relevant questions so advice can reflect the user’s circumstances.
✓
Identify signs that an urgent or safeguarding pathway may be needed.
✓
Explain local options and offer approved referral or signposting routes.
✓
Produce anonymised, aggregate insight where this has been agreed.
Support Never Stops does not…
×
Operate as a clinical system, diagnostic tool, prescribing tool or treatment system.
×
Create, hold, manage, approve or determine a service’s clinical policies, protocols or treatment decisions.
×
Replace professional assessment, judgement, emergency services, crisis teams or safeguarding professionals.
×
Give commissioners access to identifiable individual conversations.
×
Transfer or dilute a service’s accountability for clinical care, treatment decisions or local governance.
Privacy-conscious by design
Use only what is needed. Protect it throughout.
Support Never Stops is designed to provide useful support without requiring a person to create a detailed identity profile. Information use is defined by purpose, access is controlled and commissioner reporting is separated from identifiable conversation content.
Purpose limitation
Information is collected and used only for agreed support, safety, referral and reporting purposes.
Data minimisation
The experience avoids asking for unnecessary identity details and keeps questions proportionate.
Controlled access
Permissions, administrative access and operational support are limited to authorised roles.
Anonymised insight
Commissioner dashboards are designed around aggregate patterns rather than individual case surveillance.
Information lifecycle
From conversation to accountable use
01
↓
Collect proportionatelyOnly information relevant to the immediate purpose.
02
↓
Protect and restrictSecure processing and role-based access.
03
↓
Separate and anonymiseReporting data is prepared without exposing personal conversations.
04
Review and retain appropriatelyRetention, deletion and audit arrangements follow the agreed governance model.
Detailed data flows, lawful bases, retention and controller/processor responsibilities are documented for the deployment.
AI is part of the system. Accountability remains human.
Service clinical governance
The service remains solely responsible for its clinical policies, professional decisions, assessment, treatment, prescribing and clinical governance. Support Never Stops does not manage or influence these responsibilities.
Safeguarding leadership
Defines escalation principles, vulnerable-person protections and links to local safeguarding arrangements.
Information governance
Oversees privacy, data flows, access, retention, transparency and the deployment’s data-protection assessment.
Product and content governance
Support Never Stops governs its own approved information, AI behaviour, testing, versioning, releases and platform safety rules. These controls do not determine clinical policy or professional decisions.
Service and commissioner oversight
Confirms intended use, local signposting and operational ownership. Any clinical pathway, policy or decision remains outside Support Never Stops and under the service’s own governance.
Proportionate assurance
Governance evidence follows the deployment.
The final assurance set depends on how Support Never Stops is configured, what data is processed, which information and signposting pathways are enabled and how it is used alongside local services. This does not extend the platform into clinical care or clinical decision-making.
We avoid treating standards as badges. The purpose of each document, review and test is to demonstrate that risks are understood, controlled and owned.
Scope and boundary statement
Documents intended use, exclusions, responsibilities and the boundary between Support Never Stops and the service’s own clinical governance arrangements.
Privacy assessment
DPIA, UK GDPR roles, data flows, lawful use, transparency, minimisation, retention and individual rights.
Security assurance
Cyber Essentials-aligned controls, access management, vulnerability handling, independent testing, incidents and resilience.
Accessibility and inclusion
Accessible interaction, readable language, inclusive user testing and consideration of digital exclusion.
Safety is maintained through a continuing governance cycle
Define intended use
Agree users, pathways, boundaries, local context and accountable owners.
Assess platform risks
Identify safeguarding, content, privacy, security, operational and foreseeable misuse risks within the platform’s defined non-clinical scope.
Build and test controls
Test content, rules, model behaviour, pathways and foreseeable misuse.
Approve and deploy
Record decisions, residual risks, responsibilities and launch readiness.
Monitor and improve
Review feedback, incidents, performance, updates and emerging risks.
Our commitment
Transparent about capability. Clear about limits. Responsible for change.
Documented decisions and safety evidence
Named owners and human oversight
Monitoring, learning and controlled improvement
Local pathways agreed with partners
Explore the platform
and its assurance approach.
Book a tailored demonstration to review the user experience, safeguarding and safety boundaries, governance model, non-clinical scope and local deployment requirements.
Book a Demonstration