Compliance and Data Privacy FAQ

This dual approach helps to ensure high levels of accuracy, safety and educational relevance in all AI-generated responses.

Our Terms of Use (Clause 10) also reserve our right to monitor and remove content that violates the Terms or is deemed inappropriate.

Teachers and administrators can review student interactions at a granular level, including the ability to listen back to student audio recordings and leave personalised feedback on scenario performance.

Our teacher-in-the-loop approach supports transparency, enables real-time monitoring and allows educators to intervene where necessary.

Conversation and interaction data, performance data and usage analytics are collected and made available to institutions as described in Sections 3 and 4 of our Privacy Policy.

Full details are set out in Section 10 of our Privacy Policy.

Our Terms of Use explicitly prohibit jailbreaking, reverse-engineering or otherwise attempting to circumvent the AI models or security measures used on the platform (Clause 6(d)), as well as unauthorised access to servers or networks (Clause 6(f)).

Account sharing is strictly prohibited (Clause 4.3), with each account assigned to a single individual. Institutional administrators manage user access through the provisioning and revocation of licences.

We also have a formal data breach notification process, committing to notify institutions within 72 hours of becoming aware of a breach (Section 11 of our Privacy Policy).

Data controller and processor roles are clearly defined: the institution is the data controller for usage and platform data, while LanguageMate LTD acts as the data processor under a Data Processing Agreement (DPA). LanguageMate LTD is the data controller only for account registration data (Section 2.1 of our Privacy Policy).

We do not use identifiable personal data to train our AI models. Only fully anonymised data is used for this purpose (Section 5 of our Privacy Policy). Users have the right to object to anonymisation before it takes place by contacting us at [email protected].

Our platform complies with UK GDPR and the Data Protection Act 2018, with specific provisions for children and young people (Sections 6 and 9 of our Privacy Policy), including requirements for institutional and parental consent for users under 18.

Only fully anonymised data is used for these purposes (Clause 9.3 of our Terms of Use; Section 5 of our Privacy Policy).

We do not claim ownership of users' original content. To the extent that user inputs constitute copyrightable material, the licence granted in our Terms of Use allows LanguageMate to use such material only in anonymised, aggregated form for improving our service (Clause 9.3).

The platform is continuously monitored, with content moderation processes in place to review and remove inappropriate material (Clause 10 of our Terms of Use).

As generative AI models evolve over time, output quality is monitored and the platform is reviewed regularly to maintain high standards of educational relevance and safety.

Users and institutions can report violations or raise queries by contacting [email protected] (see Clause 18 of our Terms of Use and Section 15 of our Privacy Policy).

We review risks when making changes to AI models, introducing new features and updating tools, considering the impact on both educators and learners. The platform operates under clear data processing agreements with institutions, and our Terms of Use and Privacy Policy are available on our website at all times.

AI-generated responses provide personalised feedback to support skill development. LanguageMate is designed to be used alongside teacher instruction, supporting rather than replacing the educator's role.

Teachers maintain full visibility of student progress and can provide direct feedback, enabling our teacher-in-the-loop approach and ensuring that professional pedagogical judgement remains central to the learning process.

Each scenario features a different character (such as a shopkeeper, hotel receptionist or doctor), preventing the formation of emotional bonds or dependency on AI personas.

This important design decision ensures AI is used as a professional educational tool rather than a substitute for human interaction, and avoids emotionally persuasive design elements.

Detection includes sudden escalation in help-seeking or night-time usage spikes, references to mental health conditions, mentions of suicide or self-harm, use of isolation phrases and repeated refusal to end sessions.

Where distress is detected, the platform follows an appropriate response pathway, including soft signposting to age-appropriate support pages and resources and raising a safeguarding flag to the institution's designated safeguarding lead.

All response language is designed to be safe and supportive: non-validating and non-pathologising, always directing the learner to human help (such as teachers, family, peers or crisis services) and avoiding any language that suggests isolation or secrecy.

Our teacher-in-the-loop model, where educators have full oversight of student interactions and can review audio recordings, provides an additional layer of safeguarding and ensures that any concerns can be identified and addressed by qualified professionals within the institution.

Our AI does not deceive or mislead users, does not portray absolute or unjustified confidence, and does not stimulate negative emotions such as guilt or fear for motivational purposes. It does not threaten harm, loss, punishment or the withholding of benefits, nor does it apply pressure to socially conform.

There are no paid upgrade options, premium features or in-app purchases visible to individual users within the platform; all commercial matters are handled at institutional level through a separate service agreement (Clause 5.4 of our Terms of Use).

Our platform does not blend pedagogical assistance with advertisements or promotional content, does not steer users towards paid options through biased wording or layouts, and does not employ dark patterns that might deceive a user into taking unintended actions.

Our Terms of Use (Clause 5.3) include a fair use policy, and we reserve the right to restrict access for accounts exhibiting excessive usage patterns, as our platform is not designed to maximise time spent.