AI learning assistant (Aila)
Aila is a chatbot created by the University of Melbourne as a learning resource for students. Staff can customise the behaviour of Aila by creating their own ‘apps’ and can share these with others.
Interested in piloting Aila in your 2026 subjects?
We are seeking expressions of interest from subject coordinators who are interested in trialling the Aila tool in their 2026 subjects with students. The supporting teams will reach out shortly to interested staff to assist the release in your LMS subject sites.
Quick links
What is Aila?
Aila is a generative AI (GenAI) tool, that the University of Melbourne is developing as a learning resource for students. Located within subject sites in the LMS, the tool assists students with their learning by providing access to a range of chat based tools (apps) which add to the learning experience.
Aila is one of a series of initiatives aimed at capturing the benefits that GenAI can provide students and staff. The tool is classed as a beta version and is in active development. Aila uses the Large Language Model (LLM) Anthropic Claude 3 Sonnet.
Teaching and learning applications
Shaping the tool’s behaviour
The Aila global apps have been programmed to base responses on information drawn from subject content in the LMS. However with the new custom apps, subject coordinators also have the ability to use GenAI to perform a task (say, a roleplay) without being grounded in subject material. Subject coordinators can shape the tool’s responses to student queries through the material chosen as the basis for those responses, and through further ‘prompt enhancements’ within each app.
Staff can use their LMS playpen to trial the following:
- Decide which subject resources in the LMS should be part of the data set that will be used by Aila to generate responses.
- Provide additional instructions to Aila as part of the prompt enhancements. This is an advanced setting and is optional.
- Thoroughly test Aila for their subject using the View student experience function.
- Update content when subject content changes during the semester.
As with all GenAI tools powered by large language models, there is a degree of unpredictability in the tool’s behaviour.
Errors and oversight
AI systems and tools are not infallible: staff and students using Aila need to be aware of this. For every response Aila provides, the sources used to generate that response are displayed. Clicking on a source item will take you to that page or file on Canvas LMS where you can review the information directly.
More generally, staff and students should be prepared to critically review the outputs of any generative AI tool and should understand that they remain responsible for their use of these outputs.
Supporting learning and academic integrity
Aila is intended to support students in their learning rather than doing the work for students. Hence, Aila has been designed to avoid providing direct answers to assessment questions. This constraint on Aila’s behaviour has been achieved to the best of our ability, using controls over the technology that are currently available. The reliability of these constraints will be a focus of the trial phase of the learning assistant.
Students are advised that the rules around academic integrity apply equally when using Aila. If a student uses text generated by Aila in the preparation of an item for assessment, this must be appropriately acknowledged and cited in accordance with the Assessment and Results Policy (MPF1326). Unauthorised or undisclosed use of AI is a breach of the University’s student academic integrity standards, and may be subject to the penalties outlined in the Student Academic Integrity Policy (MPF1310).
University of Melbourne AI principles
Aila has been developed in line with the University’s AI principles, with an emphasis on the importance of AI literacy, academic integrity, fairness, privacy and security, and responsibility and human oversight. It is expected that the principles will likewise guide use of the learning assistant.
Access and licensing
Aila is available to all University of Melbourne staff in their Canvas LMS Playpen.
This product is currently in beta release and is being actively developed by Teaching and Learning Innovation and Business Services based on feedback from our staff and students. We are seeking expressions of interest from subject coordinators who are interested in trialling Aila tool with students in their 2026 subjects.
Students have access to Aila when the tool’s visibility is actively switched on within a subject. Teaching staff should ensure that they have tested the tool and confirmed that it is performing as expected before doing this.
Aila is housed within Spark AI, the University’s secure platform for experimenting with GenAI tools. This means that the learning assistant can be used without compromising University or student IP that is provided to the tool.
Support and resources
For technical issues, please reach out via the Teaching and Learning Innovation support request form.