Safe AI at Lancaster (SAIL) Privacy Notice

This privacy notice explains how Lancaster University processes personal data when users access Safe AI at Lancaster (SAIL), the University’s secure artificial intelligence environment. It describes what information is processed, why it is processed, how long it is stored, the lawful basis for processing, and the rights of users.

1. What data SAIL processes

SAIL processes information needed to provide, secure and govern the University's AI service for staff and postgraduate researchers.

Standard Chat is available to all University staff and postgraduate researchers.. Released workflows may be used within their published conditions without individual approval for routine tasks.

Deliberate processing of personal or sensitive information requires recorded approval for the specific use case before processing begins. Approval must specify the permitted purpose, information, users, processing arrangements and any additional safeguards. Access to approved specialist use cases will be managed by the process owner or their nominated delegate.

Permitted information classifications depend on the approved use case and validated deployment controls. Only classifications expressly recorded as approved may be processed.

The types of data that may be processed include:

  • University username, user identifier and access-group information;
  • Purpose-of-use declarations, where required;
  • Prompt text and other user inputs;
  • AI-generated outputs;
  • Uploaded files and extracted document content, where enabled;
  • Voice or dictation inputs, where enabled;
  • Technical metadata, including dates and times of use, model selection, usage data and relevant security, audit and support records; and
  • Information about individuals included by users in prompts, uploads or other submitted content.

SAIL does not automatically access student records or other University business systems unless a separate feature or integration has been formally approved and communicated to users.

Users must comply with the applicable conditions of standard Chat or their approved specialist use case and submit only information necessary for the task.

2. Special Category Data

Deliberate processing of personal or sensitive information, including special category data, requires recorded approval for the specific use case before processing begins. The accountable business or information owner, supported by the relevant University governance and security functions, must confirm the appropriate lawful basis, Article 9 condition and any additional approvals or safeguards required.

Special category personal data includes information about racial or ethnic origin, political opinions, religious or philosophical beliefs, trade-union membership, health, sex life or sexual orientation, and genetic or biometric data used to identify an individual.

Criminal-offence data, highly confidential research information, legally privileged information, commercially restricted information or other restricted data must not be submitted unless explicitly approved for the relevant use case, model, provider and processing arrangements. Any additional legal conditions or approvals required for criminal-offence data must be documented separately.

Users must remain within the approved processing conditions, submit only information necessary for the task and anonymise, pseudonymise or redact personal information wherever reasonably practicable.

Sensitive-information detection and warning controls may identify selected types of personal information in user inputs. These controls reduce risk but cannot identify or prevent all personal, confidential or sensitive information from being submitted and do not replace formal use-case approval.

3. Why SAIL processes data

Lancaster University processes personal data to:

  • Authenticate users and manage access to SAIL;
  • Provide AI-assisted chat, content generation, document analysis and other permitted features;
  • Process prompts, uploads and other user inputs to generate requested outputs;
  • Maintain security, prevent misuse and investigate suspected incidents;
  • Provide user support and resolve technical problems;
  • Maintain audit records and demonstrate compliance with University policies and legal obligations;
  • Manage service capacity, reliability, performance and costs;
  • Support service governance, planning and improvement; and
  • Provide a centrally governed alternative to unmanaged public or consumer AI services.
  • SAIL does not make solely automated decisions that have legal or similarly significant effects on individuals. Users must review and validate AI-generated outputs before using them.

Purpose records and service usage information support service operation, security, governance and planning. They will not be used for individual performance management. Access to content and records is restricted to authorised personnel with an operational need.

Approved use cases

SAIL may be used for the following approved purposes:

Student Support services: Staff use SAIL to capture confidential information for Serious Incident Reviews and student consultation cases.

4. Lawful basis for processing

The service lawful basis does not provide blanket approval for all submitted content. Use cases will be reviewed with University governance to confirm the required approvals and processing conditions before use. For information about how the University uses personal data and which lawful basis applies, please refer to the relevant Privacy Notice.

5. AI Models and cloud processing

SAIL's primary application data is stored on University-managed infrastructure. Where external AI models are used, relevant request content is transmitted to the applicable model deployment for processing.

Standard Chat uses Azure AI Foundry in Sweden . Reusable workflows use UK processing unless an exception is explicitly stated and assessed.

Deliberate processing of personal or sensitive information requires recorded approval for the specific use case, including the permitted processing arrangements.

The model deployment register records the enabled models, providers, actual processing locations, storage, retention and support-access arrangements. Any international transfers or overseas access must be assessed and covered by the appropriate contractual and data-protection safeguards.

The location of an Azure resource does not necessarily establish the actual location of AI model processing.

6. How long data is kept

Prompts, uploaded files and AI-generated outputs are retained on a rolling 30-day basis, unless a different retention period is formally approved, documented and communicated.

Users may delete individual chats where this functionality is available. Logs, access records, backups and external provider records may have different retention periods, as documented in the service retention schedule.

Deleting a chat does not necessarily result in the immediate deletion of associated logs, backups or provider records.

7. User responsibilities

Users are responsible for ensuring that they are authorised to submit information to SAIL and that their use complies with data-protection law, University policy, research-ethics requirements, contractual obligations, intellectual-property requirements and any other applicable restrictions.

Users must submit only the minimum information necessary and should anonymise, pseudonymise or redact personal information wherever reasonably practicable.

SAIL may generate inaccurate, incomplete, biased, outdated or fabricated outputs. Users must review and validate outputs before using, sharing, publishing or relying on them. SAIL must not be used as the sole basis for decisions that could materially affect an individual without appropriate human review and accountability.

SAIL is not a substitute for professional medical, legal, financial, academic, safeguarding or other specialist advice.

8. Your rights

Under UK data-protection law, individuals may have rights to:

  • be informed about how their personal data is processed;
  • request access to their personal data;
  • withdraw consent to processing where this is the lawful basis relied upon;
  • request correction of inaccurate personal data;
  • request erasure of personal data in certain circumstances;
  • request restriction of processing in certain circumstances;
  • object to processing based on legitimate interests;
  • request data portability where the legal conditions apply; and
  • complain to the Information Commissioner’s Office.

These rights are not absolute and may apply only in specific circumstances.

Users may view and, where available, delete their own chat history within SAIL. Requests relating to information held in service records, audit logs, backups or other technical records will be assessed through the University’s established data-subject-rights process.

For further information on your rights regarding your personal data, and how to exercise them, please see the University’s Rights of the Data Subject webpage.

9. Contact details and complaints

Lancaster University’s designated Data Protection Officer is:

Mike Abbotts, Information Governance Manager
Email: information-governance@lancaster.ac.uk

Contact Mike if you have any concerns or complaints about this Notice or about the way your personal data is being used. You also have the right to complain to the Information Commissioner’s Office (ICO). Details of how to raise a complaint can be found on the ICO website.

Changes to this policy

We may change this privacy policy as the service for SAIL or Data Protection legislation changes.