Cogniti is an education-first GenAI platform developed by the University of Sydney. It has been designed to let teachers build custom chatbot agents that can be given specific instructions, and specific resources, to assist student learning in context-sensitive ways.
Unlike many generic GenAI tools, Cogniti is purpose built for education, by educators. It supports equitable student access to advanced models, and options to anchor responses in reference to your own course materials and context.
The University of Sydney group behind the platform, driven by the centrality of human teachers and the importance of pedagogy, describe it as Steerable, Resourced, Equitable, and Insightful.
| Create AI agents and control how they will interact with students. | Provide your AI agents with web pages, files, and other resources to answer more accurately. | Provide all students with the same access to powerful AI agents. | Understand how your students are interacting with AI agents, to inform your teaching. |
Amended from information available at https://cogniti.ai/
Following a successful trial across more than a dozen courses, access to Cogniti is being rolled out from Trimester/Semester 1 2026. This page has been developed to outline the process for accessing the platform, describe the expectations for academics who elect to use the platform in 2026, and describe options for help and support.
Throughout the trial stage, academics using the platform noted the relative ease of set up and intuitive, user-friendly interface. Based on this feedback, implementation of Cogniti assistants will be led by individual educators with support provided by LDTI staff.
Top Tips for Using Cogniti
Unlike many other general purpose AI options, Cogniti has been developed with an “education first” purpose. When thinking about how you may use a Cogniti agent in your course(s), always start by defining the educational purpose for using the tool. For Cogniti to be most effective, it needs to be aligned to the purpose of your course.
To get the most benefit from the platform, it is recommended that you:
- Give students the context: Explain the purpose of the Cogniti agent, and why you have chosen to use it.
- Model how to interact with the agent: Consider an in class or recorded demo of the agent in action.
- Emphasise the value for learning: Position the tool as a learning opportunity, not a source of “easy answers”.
- Invite reflection and feedback: Gathering student feedback also helps you refine the agent and its integration in your unit
Adapted from the University of Sydney’s Four strategies for implementing custom AIs that help students learn, not outsource (September 2025).
Cogniti access will be available to all 2026 course coordinators who have engaged with the initial training opportunities.
To help us tailor that training and support, we are asking all academics to register their interest using this form:
Cogniti Registration Form
This will include providing some basic details of your idea(s) for using Cogniti (if known). Please complete the form by 12 January 2026 to be included in the first training sessions.
“Overall, the process of creating Mrs Lin was remarkably efficient, requiring approximately 2.5 hours of work…..The platform is very intuitive and user-friendly for both agent creators (i.e. educators) and end-users (i.e. learners), providing clear directions on each page.
We expect that academic staff would be able to create Cogniti agents without professional staff support, if provided with brief guidance through a video guide or step-by-step guide.”
Learning Design and Teaching Innovation will run an online introductory workshop, before the end of January 2026, for staff who have registered using the form above. While the session will be recorded and shared, attendance during the session is preferred.
Prior to being granted access to create Cogniti agents, staff will need to complete a declaration confirming that they have participated in the introductory training, and aware of other requirements (e.g. transparent disclosure to students, participation in feedback processes, etc) for use.
Additional support is available through the guides below, and by contacting LDTI for assistance. Additional support resources will be added in early 2026.
Throughout the first year of use at our institution, academic staff who choose to use Cogniti in their course(s) will be required to engage in some low touch feedback opportunities.
These will help improve the use of the platform across the institution and inform decisions about continued use beyond 2026.
Cogniti users will be required to:
- Consent to 2-5 standard questions being included in their Course Experience Survey. Results from these surveys will be de-identified and used only to gauge student sentiment on Cogniti. LDTI will work directly with Strategy, Planning and Performance to add these questions to the relevant 2026 CES surveys. Course Coordinators will also have access to their own results.
- Consider opportunities to utilise LDTI’s Mid-Semester Student Feedback process to source actionable student feedback to feed directly into your current teaching or future iterations of the course.
- Complete a brief end of term questionnaire (administered by LDTI) on their views of Cogniti in their course(s).
Sector-wide and internal collaboration strengthens collective expertise.
Effective GenAI practice grows through open exchange of insights among colleagues, universities, professional bodies and industry, and through focused dialogue within each academic field to address its unique standards and challenges.
By sharing knowledge at all levels, we can refine our strategies and stay aligned with emerging expectations.
To help our academic community learn more about the potential benefits and risks of GenAI in education, we hope that users will take the opportunity to participate in opportunities to share their successes (and failures!) using Cognit in 2026.
This may include
- Presenting at the Educator Network (tEN) events such as showcases and learning lunchboxes.
- Collaborating with LDTI on AI in Action Bite Size Teaching Sessions.
- Appearing in future instructional and support videos.
- Share your experiences via the AI Community of Practice
- Considering opportunities to participate in external events such as the annual University of Sydney Cogniti Mini Symposium.
The platform provides flexibility to create agents and assistants that can be applied to support a variety of educational use cases.
Some of the more common applications include:
Course Tutor: Agents built to support students’ engagement with course content.Some examples:
|
Feedback and Assessment Preparation: Agents that provide opportunities to access AI feedback or engage in practice assessment.Some examples:
|
Digital Clients: Agents built to facilitate students’ engagement in digital experiences through engagement with virtual patients, clients, students, etc.Some examples:
|
Course Assistant: Agents built answer questions about the course (assessment dues dates, contact info, etc).Some examples:
|
Course/Learning Design Support: Agents built to assist and augment design processes.Some examples:
|
Specific Learning Activities: Agents designed to allow students to complete specific learning activities.Some examples:
|
Frequently Asked Questions
LDTI can provide guidance and assistance. However, as with any educational technology or learning materials, Course Coordinators are ultimately responsible for how it is used and the quality of their teaching and learning experiences.
Additional information is available under Course Co-ordinator Responsibilities in the Course Design and Management Manual.
The use of GenAI is a relatively topic. However, as adoption continues to increase, more information and studies are becoming available. Some further reading is included below:
- Belkina, M., Daniel, S., Nikolic, S., Haque, R., Lyden, S., Neal, P., Grundy, S., & Hassan, G. M. (2025). Implementing generative AI (GenAI) in higher education: A systematic review of case studies. Computers and Education: Artificial Intelligence, 8, 100407. https://doi.org/10.1016/j.caeai.2025.100407
- Kestin, G., Miller, K., Klales, A. et al. (2025). AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting. Sci Rep 15, 17458 https://doi.org/10.1038/s41598-025-97652-6
- Flenady, G., & Sparrow, R. (2025). Cut the bullshit: why GenAI systems are neither collaborators nor tutors. Teaching in Higher Education, 1–10. https://doi-org.ezproxy.newcastle.edu.au/10.1080/13562517.2025.2497263
- Tibbs, A. (2025). Building capacity for AI-powered learning: A collaborative trial of conversational agents in higher education. 2025 ASCILITE Conference Proceedings. https://doi.org/10.65106/apubs.2025.2647
Several case studies and examples are available on the Academic Integrity, Artificial Intelligence, and Standards Working Group site at HIGHLIGHTED RESOURCE: Cogniti - AI Agents Created by Teachers
Refer to information on the Cogniti webpage. Cogniti on Microsoft Marketplace is governed by the agreements established during your Microsoft Marketplace purchase. This deployment option runs Cogniti within your organisation’s dedicated Azure tenancy.
- Data is encrypted in transit and at rest using industry-standard technologies
- Data isolation within your Azure tenancy
- Your data is never used to train AI models unless you choose to do so yourself
Cogniti is one of a new generation of software that uses generative AI to augment human intelligence and work. See more in this University of Sydney article.
The University of Sydney provides detailed information on the Cogniti webpage.
Frequently Asked Questions
LDTI can provide guidance and assistance. However, as with any educational technology or learning materials, Course Coordinators are ultimately responsible for how it is used and the quality of their teaching and learning experiences.
Additional information is available under Course Co-ordinator Responsibilities in the Course Design and Management Manual.
The use of GenAI is a relatively topic. However, as adoption continues to increase, more information and studies are becoming available. Some further reading is included below:
- Belkina, M., Daniel, S., Nikolic, S., Haque, R., Lyden, S., Neal, P., Grundy, S., & Hassan, G. M. (2025). Implementing generative AI (GenAI) in higher education: A systematic review of case studies. Computers and Education: Artificial Intelligence, 8, 100407. https://doi.org/10.1016/j.caeai.2025.100407
- Kestin, G., Miller, K., Klales, A. et al. (2025). AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting. Sci Rep 15, 17458 https://doi.org/10.1038/s41598-025-97652-6
- Flenady, G., & Sparrow, R. (2025). Cut the bullshit: why GenAI systems are neither collaborators nor tutors. Teaching in Higher Education, 1–10. https://doi-org.ezproxy.newcastle.edu.au/10.1080/13562517.2025.2497263
- Tibbs, A. (2025). Building capacity for AI-powered learning: A collaborative trial of conversational agents in higher education. 2025 ASCILITE Conference Proceedings. https://doi.org/10.65106/apubs.2025.2647
Several case studies and examples are available on the Academic Integrity, Artificial Intelligence, and Standards Working Group site at HIGHLIGHTED RESOURCE: Cogniti - AI Agents Created by Teachers
Refer to information on the Cogniti webpage. Cogniti on Microsoft Marketplace is governed by the agreements established during your Microsoft Marketplace purchase. This deployment option runs Cogniti within your organisation’s dedicated Azure tenancy.
- Data is encrypted in transit and at rest using industry-standard technologies
- Data isolation within your Azure tenancy
- Your data is never used to train AI models unless you choose to do so yourself
Cogniti is one of a new generation of software that uses generative AI to augment human intelligence and work. See more in this University of Sydney article.
The University of Sydney provides detailed information on the Cogniti webpage.
Resources
For detailed help documentation, refer to the Cogniti Website. Additional UON support resources and videos will be added in 2026.
Cogniti is an award-winning platform built by educators that lets you create course-aware AI assistants you control and embed in Canvas to support your students. It runs on trusted platforms, uses your own materials to answer, supporting institutional governance and privacy needs.
Additional detailed information is available on the Cogniti Help Page.
Creating a basic Cogniti agent is a relatively straight forward process based around the use of a “system message”.
See more on the Cogniti website at:
The system message (or prompt) is the set of instructions, written in natural language, that direct your agent who to act. No coding or other technical expertise is required to write a system message
See more on the Cogniti website at:
The performance of your agent can be improved through the inclusion of suitable resources. Your agent can reference these resources, leading to improved accuracy and anchoring responses within your course context.
See more on the Cogniti website at:
Once you create an agent in Cogniti, you can embed this in a Canvas page to make it easier and more seamless for students to use your agent. This way, they don’t need to create an account, and can use your agents right where they are engaging with learning.
See more on the Cogniti website at:
Course Coordinators are ultimately responsible for how Cogniti is used and the quality of their teaching and learning experiences.
Before being able to deploy an agent, CCs will need to confirm that they have reviewed the available support materials, tested and refined their agent, and are comfortable using it in their teaching.
After first drafting your system message and adding any relevant documents, you should iteratively test to investigate how your agent responds, refine and adjust. You should also continue to monitor interactions and outputs after deployment.
See more on the Cogniti website at:
The University of Newcastle acknowledges the traditional custodians of the lands within our footprint areas: Awabakal, Darkinjung, Biripai, Worimi, Wonnarua, and Eora Nations. We also pay respect to the wisdom of our Elders past and present.