SkyOwl is a research prototype developed through the University of Melbourne project Safe by Design: Building Human-Centred Digital Flight Assistants (Human Research Ethics ID 31176). This Privacy Policy explains what information we collect through the SkyOwl prototype, how we use and protect it, and the choices you have.
01Who we are
The prototype is operated by the SkyOwl research team at the University of Melbourne. The Responsible Researcher and data controller is A/Prof Fabio Mattioli, School of Social and Political Sciences, University of Melbourne, Parkville VIC 3010, Australia. The project is funded by the University of Melbourne, the Australian Government Department of Education, and AvSoft Pty Ltd.
02Information we collect
We collect only what we need to run the prototype and conduct the research:
- Account information — your name and email address, and (if you provide them) your role and organisation, so we can give you access and stay in touch about the project.
- Your interactions with the prototype — the messages, prompts, scenarios, and responses you exchange with SkyOwl, whether through the web interface or a messaging channel such as WhatsApp.
- Session content — where you have separately consented during a co-design or testing session, audio or video of that session. You can ask us to alter or obscure your voice or face.
- Technical and usage data — basic device, log, and usage information (for example, time of access and features used) needed to operate and improve the prototype.
You are in control of the information you share with the prototype. Please avoid entering real names of individuals or organisations, or operationally sensitive information, unless it is necessary for the test.
03How we use your information
- To provide you with access to the prototype and operate it during the testing phase.
- To conduct research on how AI can support aviation training in non-technical skills (decision-making, communication, situational awareness, threat-and-error management).
- To develop, evaluate, and improve SkyOwl, including a future product of a commercial nature.
- To train and fine-tune AI models using de-identified data, under the supervision of the research team.
- To communicate with you about the project, its results, and related activities.
04Legal basis — your consent
We process your information on the basis of your consent. Participation is entirely voluntary. You may withdraw at any time without explanation or prejudice, and you can ask us to limit the use of, or delete, your unprocessed data up to two weeks after it is collected. See Account & Data Deletion.
05How your information is shared
We do not sell your personal information and we do not use it for advertising. Your information may be accessible to:
- The research team — only members of the SkyOwl / University of Melbourne research team can access raw data.
- AI processing (Anthropic via Amazon Bedrock) — to generate responses, your interactions are processed by Anthropic AI models running on Amazon Web Services (AWS Bedrock) under a zero-data-retention configuration. This means your interactions are processed transiently to produce a response and are not retained by the AI provider. This processing takes place in AWS data centres located in the United States and Australia.
- Messaging platforms — if you choose to interact with the prototype over a channel such as WhatsApp, that platform's own privacy terms also apply to messages sent through it.
- A future commercial partner — we plan to lease a dataset containing de-identified data to a spin-off partly owned by the University, which will use it to build a commercial training assistant. Access to this database is regulated by A/Prof Mattioli to minimise the risk of data breaches.
- Where required by law — the confidentiality of your information is safeguarded subject to any legal requirements.
06Storage, location, and retention
Short-term interactions — while you use the prototype, your interactions are processed on cloud servers (AWS) located in the United States and Australia, under the zero-data-retention terms described above.
Long-term storage — data that we retain for the research is stored securely on University of Melbourne servers. De-identified research data is held in a secure databank in perpetuity for research and development purposes.
You can request deletion or restriction of your unprocessed data up to two weeks after it is collected. After that point, and once data has been de-identified and incorporated into the research or model training, it can no longer be individually withdrawn.
07Security
We use secure University storage and access controls, and we work only with de-identified data wherever possible. No system is perfectly secure, and we cannot guarantee the security of data while it is being processed by third-party providers.
08Your rights and choices
- Withdraw from the project at any time.
- Access or correct the personal information we hold about you.
- Delete your account and request removal of your unprocessed data — see Account & Data Deletion.
- Alter or obscure your voice or face in any recording.
09Children
SkyOwl is intended for aviation professionals and invited participants aged 18 or over. It is not directed at, and we do not knowingly collect information from, anyone under 18.
10Changes to this policy
We may update this policy as the project develops. We will post the revised version here with a new effective date, and where changes are significant we will notify participants by email.
A/Prof Fabio Mattioli, Responsible Researcher
Email: fabio.mattioli@unimelb.edu.au · Tel: +61 3 8344 1291
Research Integrity Administrator, Office of Research Ethics and Integrity, University of Melbourne, VIC 3010
Tel: +61 3 8344 1376 · Email: research-integrity@unimelb.edu.au
Please quote Ethics ID 31176 and the project name.