Privacy Policy

What data does Oxford Language Technologies Ltd [OLT] collect and how do we use it?

What personal data do we collect?
Except where necessary to provide our service, OLT keeps personal identifying data to an absolute minimum.
We use Microsoft Azure AD B2C to manage user accounts.
  • You log in using an existing account which you have with a major internet provider, such as Microsoft, Google, Apple or Amazon.
  • We never see your password and do not request any sensitive or restricted data from the account provider. Audit logs of login attempts are kept by Microsoft for 7 days.
  • Once Microsoft has verified your login credentials, you are assigned an ID number unique to Omniloquent Language Primer and which cannot be used to link the data collected by OLT to any other website or app which you log into.
What other data do we collect?
We collect your responses to the tasks in the app, including:
  • The time and date.
  • What you type.
  • What you say while pressing the button.
  • Speech recognition results and pronunciation evaluation.
  • How long the trial lasted and how long it took you to respond.
  • Whether you got the trial correct.
  • Running statistics on what you have seen, how many times you have seen it, how many times you got it correct &c.
On what basis is this data collected and processed?
Under UK GDPR, the data collected by OLT is collected and processed on a 'legitimate interests' basis, compliant with Article 6(1)(f).
What is our legitimate interest?
  • Why do you want to process the data – what are you trying to achieve?
    The data is processed in order to provide the best possible language learning tools.
  • Who benefits from the processing?
    The user of the Omniloquent Language Primer benefits directly from improvements to the tools derived from processing the data.
  • Are there any wider public benefits to the processing?
    Society benefits from improved ability for people to communicate with each other, and from improved scientific understanding of language and language learning.
  • How important are those benefits?
    The study of language is well established and funded within academia globally.
    In AI, Large Language Models (LLM) are currently poised to effect significant societal changes and challenges.
  • Would your use of the data be unethical or unlawful in any way?
    No.
Is the processing necessary?
  • Does this processing actually help to further that interest?
    Yes. The personal data collected is necessary to offer our service and carry out research.
  • Is it a reasonable way to go about it? Is there another less intrusive way to achieve the same result?
    The intrusion on people's privacy is minimal.
What is the impact of our processing on the individual (the balancing test)?
  • Is any of the data particularly sensitive or private?
    The personal data collected is minimal. Further, within the Omniloquent Language Primer web site / app, the data can be considered pseudonymous as it can only be linked to an individual's name or email address through the account details stored in Microsoft Azure AD B2C, and to which researchers have no access.
    In some circumstances, audio recordings of speech can be considered special category data, as the recording may be used to reveal racial or ethnic origin, which is protected under Article 9. We do not believe that to apply in this case as we are not processing to data to make any such inference, and we are not using the recordings as biometric data to identify the person.
  • Would people expect you to use their data in this way?
    Yes.
  • Are you happy to explain it to them?
    OLT are committed to the principals of Open Science and will publish and promote the research generated though the use of this data.
  • Are some people likely to object or find it intrusive?
    No.
  • What is the possible impact on the individual?
    We have identified no negative impacts on the individual.
  • Are you processing children’s data?
    Yes.
  • Are any of the individuals vulnerable in any other way?
    No.
  • Can you adopt any safeguards to minimise the impact?
    All data is looked after to the highest standard as detailed below, and is used anonymously or pseudonymously wherever possible.
How is this data looked after?
  • The data is stored in the Microsoft Azure cloud, in the UK.
  • All data is encrypted at rest and when transmitted over the network.
  • Researchers can only access the data via the Omniloquent Language Primer website, and do not have direct access to the underlying databases and storage.
  • Only anonymous data can only be exported.
  • Where Google Cloud services are used, no data is stored within Google's systems. Data is encrypted in flight and processed only in-memory.
Who is this data shared with?
  • Data may be shared with accredited researchers. These researchers will have their own ethics approval and data policy, approved by their academic institution. All scientific research carried out by OLT or through OLT's technology must abide by the principals of Open Science
  • All data collected by OLT may also be used by OLT's researchers and developers in order to monitor the operation of the website, produce aggregate statistics, improve the operation or design of the web site / app, to improve OLT's course materials, or to further scientific understanding of language and language learning.
  • What you write / say can be passed securely to AI service providers who provide speech recognition and pronunciation assessment, and to Large Language Models (LLM) such as ChatGPT to provide personalised feedback and assistance.
  • OLT may collaborate with AI service providers to improve and fine tune their services, or to develop new services.
  • Any government or law enforcement agency with the legal right to view the data.
Can I view my history or listen to my voice recordings?
  • A high level log of every task you carry out in the app is available on the History page.
  • On that page you can hear each recording, and delete individual recordings or entire history entries.
How is Google user data used?
  • If you log in using a Google account, basic profile information is passed to Azure AD B2C. OLT does not use this Google user data for anything except account management, and does not share the data except as required for account management.
What personal data do we collect?
All experiments where you are enter a code on the home page are anonymous.
  • OLT does not collect any identifying information, such as names, email addresses, IP addresses, phone numbers &c.
  • For single session experiments OLT does not collect the code you enter to start the experiment, making it impossible to link your responses with external information.
  • For multiple session experiments OLT uses the code you enter to start the experiment to link the sessions together, however the code is never visible to any researcher. Once the experiment is complete the code is deleted, making it impossible to link your responses with external information.
What other data do we collect?
We collect your responses to the trials in the experiment, including:
  • The time and date.
  • What you type.
  • What you say while pressing the button.
  • Speech recognition results and pronounciation evaluation.
  • How long the trial lasted and how long it took you to respond.
  • Whether you got the trial correct.
  • Running statistics on what you have seen, how many times you have seen it, how many times you got it correct &c.
How is this data looked after?
  • The data is stored in the Microsoft Azure cloud, in the UK.
  • All data is encrypted at rest and when transmitted over the network.
  • Researchers can only access the data via the Omniloquent Language Primer website, and do not have direct access to the underlying databases and storage.
  • Only anonymous data can only be exported.
  • Where Google Cloud services are used, no data is stored within Google's systems. Data is encrypted in flight and processed only in-memory.
Who is this data shared with?
  • Experiment data is shared securely and anonymously with the researcher(s) who are running the experiment. The researchers will have their own ethics approval and data policy, approved by their academic institution. All scientific research carried out by OLT or through OLT's technology must abide by the principals of Open Science
  • All data collected by OLT may also be used by OLT's researchers and developers in order to monitor the operation of the website, produce aggregate statistics, improve the operation or design of the web site / app, to improve OLT's course materials, or to further scientific understanding of language and language learning.
  • What you write / say can be passed securely to AI service providers who provide speech recognition and pronunciation assessment, and to Large Language Models (LLM) such as ChatGPT to provide personalised feedback and assistance.
  • OLT may collaborate with AI service providers to improve and fine tune their services, or to develop new services.
  • Any government or law enforcement agency with the legal right to view the data.
What Cookies do we store?
Session Cookies are used to identify the currently logged in user or experiment participant, and to ensure that all requests are sent to the correct server in our server pool.
If you have access to multiple languages, a cookie is used to remember which language you are currently learning.
We use Azure Application Insights to monitor the availablilty and performance of the app.
All Cookies are strictly necessary to provide our language learning services.