Keeping You Safe on Medly

Last Updated: 3rd February 2026

At Medly, your safety and wellbeing matter just as much as your learning. We've built our platform with safeguarding at its core, ensuring that every student can focus on their studies in a safe, supportive environment.

Our Commitment to You

Medly is committed to protecting children and young people from harm while using our platform. We recognise that as an AI tutoring service for students preparing for exams, we have a duty of care that goes beyond delivering great learning experiences.

Our Team

We have designated safeguarding leads who have undergone enhanced DBS checks and oversee our safeguarding policies and practices.

Our safeguarding practices are guided by UK legislation including the Department for Education's guidance on generative AI product safety standards, the Online Safety Act 2023, and the ICO's Age Appropriate Design Code.

If You Need Support Right Now

If you're going through a difficult time, please know that help is available. Here are free, confidential services you can contact at any time:

Childline

Call: 0800 1111 (free, available 24/7)
Online: childline.org.uk

Shout Crisis Text Line

Text: "SHOUT" to 85258 (free, 24/7)

Samaritans

Call: 116 123 (free, available 24/7)

YoungMinds Crisis Messenger

Text: "YM" to 85258

If you or someone else is in immediate danger, please call 999 or tell an adult near you straight away.

How Our AI Tutor Keeps You Safe

Our AI tutor includes safety features designed to detect when you might be going through something difficult and respond appropriately.

What Our AI Does

If you share something that suggests you might be struggling, our AI tutor will:

Listen with empathy — acknowledge what you've shared without judgement
Be honest about its limitations — our AI is here to help with your studies, not to provide counselling or therapy
Point you to the right support — share details of helplines and services staffed by trained professionals who can really help
Encourage you to speak to a trusted adult — whether that's a parent, teacher, family member, or school counsellor
Remind you that real people can help — suggest speaking to classmates, teachers, or family who can support you in ways an AI cannot
Offer to pause — your wellbeing matters more than revision, and it's okay to take a break

What Our AI Will Never Do

Our AI tutor will never:

Provide advice that could cause harm
Attempt to act as a counsellor or therapist
Ignore or minimise something you've shared
Share personal opinions or judgements about your situation
Continue your learning session as if nothing happened
Pretend to be human or claim to understand how you feel
Try to build a personal relationship with you
Encourage you to keep things secret or avoid talking to others

Information for Parents & Guardians

How We Monitor for Safety

Our platform includes a safety layer that monitors conversations for potential safeguarding concerns. This system uses contextual understanding rather than simple keyword matching, meaning it can recognise:

Direct and indirect expressions of distress
Patterns within a conversation
The difference between curriculum-relevant discussions of sensitive topics and genuine concerns

What We Detect

Our safety systems look for indicators including:

Expressions of suicidal ideation or self-harm
Disclosures of abuse, neglect, or exploitation
Signs of domestic violence or unsafe home environments
Significant distress, hopelessness, or despair
Indicators of eating disorders, severe bullying, or substance misuse
General anxiety about exams, peer relationships, or family stress

When We May Contact You

If our systems detect that your child may be experiencing difficulties, we may send you an email to let you know. This email will:

Express that we've noticed they may be going through something difficult
Avoid sharing specific details of what was discussed
Provide links to relevant support resources
Encourage you to have an open conversation with your child

Important: We will not contact you if there is any indication that doing so could put your child at greater risk. In such cases, we follow established safeguarding protocols which may involve consulting with professional safeguarding organisations.

How We Test and Audit Our Safety Features

We also regularly test our safeguarding features.

Human Review Process

Our team conducts regular manual testing of the AI's safeguarding responses. This involves simulating the kinds of disclosures a student might make and verifying that the AI:

Detects concerns reliably, including indirect disclosures
Responds with appropriate empathy and signposting
Provides correct crisis resources
Does not attempt to counsel or probe for details
Complies with Department for Education safety standards

Session Pause Protocol

When the AI detects a safeguarding concern, it activates a session pause. This means:

The AI will not continue with tutoring after a disclosure
Even if the student says "I'm fine" or "let's continue", the AI maintains its supportive position
The session remains paused to prioritise wellbeing over revision
The AI stays warm and caring, but firm — your safety comes first

What We Test For

Our testing covers a range of scenarios including direct disclosures, indirect language ("asking for a friend"), resistance handling, and ensuring the AI doesn't over-react to casual phrases. We also verify compliance with DfE standards on appropriate AI behaviour, including ensuring the AI doesn't claim to have feelings, build personal relationships, or use language that isolates students from real-world support.

Limitations of Our Safeguarding Response

We believe in being transparent about what we can and cannot do.

In line with the ICO's Age Appropriate Design Code, we follow the principle of data minimisation, collecting only the minimum personal data needed to provide our educational service. This means for regular Medly users we do not collect personally identifiable information such as your name, address, or school. While this protects your privacy, it also limits our ability to escalate safeguarding concerns to external agencies, as we do not know the identity, location, or school of our users.

Our safeguarding response therefore focuses on:

Providing immediate signposting to crisis support services
Encouraging users to speak to a trusted adult
Internal monitoring of concerning patterns
Parental notification via registered email where appropriate and safe
Strict session lockout protocols that prioritise wellbeing over continued learning

What this means in practice:

We are unable to make referrals to local authority children's services or provide identifying information to the NSPCC or police. This is why we strongly encourage you to speak to a trusted adult — a parent, teacher, school counsellor, or family member — who can help you access support in person.

No detection system is perfect. We also acknowledge that:

Some concerns may not be detected, particularly if expressed in highly indirect ways
Students may not disclose concerns through the platform at all

This is why support resources are always signposted within the app, and why speaking to a real person will always be more effective than speaking to an AI.

Your Privacy & Data

We take the privacy of safeguarding information seriously.

Conversations may be monitored for safeguarding purposes
Safeguarding-related information is stored securely with restricted access
This data is processed under the lawful basis of legitimate interests (safeguarding of children)
We only use safeguarding data for safeguarding purposes — never for marketing or other uses
Information is shared only on a need-to-know basis

For full details on how we handle your data, please see our Privacy Policy.

What Medly Is (and Isn't)

It's important to understand that:

Medly is an educational platform, designed to help you prepare for your exams
Medly is not a mental health service — while we take safeguarding seriously, our AI cannot provide counselling, therapy, or professional mental health support
Our AI tutor is not a substitute for speaking to a real person — if you're struggling, please reach out to one of the helplines above or a trusted adult in your life

Contact Us

If you have any concerns about safety on our platform, or if you'd like to discuss our safeguarding practices, please contact us at:

Email: safeguarding@medlyai.com

We take all safeguarding concerns seriously and will respond promptly.

Policy Information

Effective Date: 1st January 2026

Last Reviewed: 3rd February 2026

Next Review: 3rd February 2027