Durmah Legal Studies
Durmah Learning Pte Ltd

Responsible Study and Student Rights

Durmah exists to help law students develop their own understanding and judgment. Our tutoring approach centres on explaining concepts, asking useful questions, organising study and encouraging students to check sources. Students choose their goals and remain in control of their work.

We are an independent service operated by Durmah Learning Pte Ltd, Singapore. We have no university affiliation or endorsement. A university name describes the intended audience or study context; it does not establish approval, source rights or complete syllabus coverage.

What students can do

Learn ideas and explain them independently. Copyright protects original expression, rather than ideas, methods or principles themselves. Studying a doctrine does not give a university ownership of that doctrine. Its slides, diagrams, recorded lectures and written explanations may nevertheless be protected. Independently explaining a concept is different from reproducing a protected explanation. WIPO copyright guidance.

Use their own original work. Students generally own copyright in original notes and writing they create, subject to relevant agreements and any third-party content included in them. Paying tuition or receiving a document does not transfer its copyright or automatically permit every subsequent use. A university, lecturer, publisher or another rights holder may own different parts. Bodleian guidance for students.

Rely on a licence or an applicable exception. UK law permits fair dealing for private study; permission from the owner is not required when the legal conditions are met. Fairness is fact-specific, including the amount needed and impact on the original work. There is no universal percentage or blanket permission to upload an entire course collection. IPO exceptions guidance.

Some copyright exceptions expressly prevent contracts from overriding qualifying uses. That does not turn every upload into an exception or resolve separate confidentiality, data-protection or assessment requirements. CDPA section 29(4B).

Private tutoring, cloud storage and AI

The tutor analogy explains our educational purpose: support the learner's understanding without replacing their reasoning. It is not a separate legal exemption. A human tutor and a software service must each consider the acts they perform and the rights needed for them.

Electronic storage can count as copying under UK copyright law, even without public posting. Uploading, extracting text and passing it to a model may involve distinct copies. Calling a workspace private or an AI a secondary intelligence does not settle whether those acts are permitted. CDPA sections 16–17.

Copying is not automatically infringement. A student can ordinarily authorise storage and retrieval of their own original writing; licensed material and applicable exceptions can also permit particular uses. Word documents, private cloud storage, calendars and AI assistance should be assessed by the actual operation and rights involved. Our comparison with Microsoft study tools explains why legitimate cloud use supports this model, while storage, inference, training and assessment permission remain distinct.

There is also a limited exception for certain temporary technical copies. The PRCA litigation concerned screen and browser-cache copies during internet browsing; it does not establish blanket permission for retained AI uploads or commercial course libraries. Deleting a file later does not, by itself, establish the exception. UK Supreme Court case and subsequent CJEU judgment.

Private, individual use and avoiding redistribution are meaningful design choices. They reduce exposure and may matter to the legal analysis, but do not answer every copyright question. A paid service is not automatically unlawful; a student's private-study purpose does not automatically authorise every act of its commercial provider either.

The UK text-and-data-analysis exception concerns lawful access and non-commercial research, with additional conditions. It is not a general commercial AI licence. The government's March 2026 copyright-and-AI report discusses policy and legal issues; publishing that report did not itself create a new permission for our service. IPO guidance; government report.

Individual study support and model training

Your study material is for your learning. Durmah's product rule is to process permitted material for the individual student's requested study assistance and the necessary service operations described in our privacy policy. That permission does not include using the material to train or fine-tune AI models, build shared model-improvement datasets, or provide another student's answers.

Generating a response from your prompt and permitted material is called inference. It is different from training or fine-tuning a model. The AI may use its existing knowledge and authorised reference sources to help explain your material; it is not limited to repeating your document. Another student's private work must not supply your answer.

This requirement applies to Durmah and its AI providers. Before release, provider contracts and settings must be reviewed for this restricted use of student inputs, outputs and derived content. These reviews are incomplete, so external AI and new academic uploads remain paused. No training is different from no storage or no operational access. Read our privacy policy and current availability.

For example, where the necessary permissions exist and a feature has been released, a student could request an explanation of a difficult passage. The passage supplies context for that student's response. Our rule does not permit turning that passage, the response or a derived summary into shared training or evaluation data. Keeping a student's permitted history or private search index is a separate storage purpose with its own access and retention requirements; it is not automatically model training.

The ICO distinguishes AI development from deployment and requires the purpose and lawful basis of each personal-data processing operation to be considered separately. Its guidance is under review following the Data (Use and Access) Act. This supports a clear purpose distinction; it does not create a copyright exception or permission to use AI in an assessment. The original model's training data is also a separate question from Durmah's handling of student submissions.

How we apply these principles

  • Start with the learner. Use your own explanation of a concept or the minimum factual context needed. A topic such as “consideration in contract law, week 3” can help plan study without importing a handbook. Do not copy a protected collection merely by changing its wording.
  • Check the actual material and operation. Ownership, a suitable licence, permission or an applicable exception must cover the intended use. Access to a course portal is not enough by itself. Our release policy requires documented permission or a suitable licence for third-party uploads; it does not currently offer an automatic fair-dealing approval route.
  • Keep individual content separate. Student uploads are not used to create shared university packs or public learning resources. Account access controls separate student records. Operational access, providers and international processing are explained in our privacy policy.
  • Be precise about AI processing. A private account is not the same as device-only storage. Provider contracts, retention, training settings and transfers remain under review. We do not claim universal zero retention, no training or UK-only processing.
  • Respect assessment instructions. Copyright permission does not establish academic permission. Follow the current rules for the programme, module and assessment, including restrictions on planning, feedback and editing. Acknowledging AI assistance cannot override a prohibition.
  • Support independent judgment. The planned initial coach asks diagnostic questions about a student's existing reasoning. It is not an official marker, an authorship certificate, legal advice or a guarantee of better grades. Check authorities, jurisdiction and currency yourself.

Read the student guide with practical examples, content permissions and AI use policy.

Evidence, corrections and review

Publishing a policy does not establish compliance on its own. Our current release controls keep AI and uploads closed while provider, privacy, material-rights and task-permission reviews remain incomplete. Reopening requires evidence for the particular feature and material; a student declaration is not a substitute.

Source review date: 18 September 2026. This page separates general legal principles, Durmah's operating choices and work still under review. It is an educational explanation, not a legal opinion about a particular file or a certification by a university, regulator or court. Applicable laws and institutional instructions can change; the date records this review, not continuous legal monitoring.

Students, educators and rights holders can report an inaccurate claim or material concern to admin@durmah.ai. Please identify the page, material and concern, and link to the relevant source where possible. Do not email confidential course documents or student records. We can assess a concern without involving a student in publicity or asking them to defend the service.

Legal sources and scope

The copyright discussion concerns UK law. Durmah's international operations and each student's circumstances may raise additional obligations. Our choice to require reviewed permissions is an operating safeguard, not a statement that all other private study is unlawful.