Student
Governed learning assistant
Support shaped by the course, assignment, and permitted capabilities.
Why RedAPPL
Universities shouldn’t react to AI.
They should shape it.
AI is becoming part of how students learn, faculty teach, and institutions operate. The university should define how it works.
RedAPPL is designed to give universities a control layer for AI across the institution—down to the course and assignment.
AI-native by design. Governed by the university.
Model choice
- OpenAI
- Anthropic
- Microsoft
- Local models
controlled through
RedAPPLThe university’s control layer
governed across
- Institution
- Department
- Course
- Assignment
01The shift
AI isn’t another classroom tool. It’s becoming infrastructure.
Institutions increasingly have to coordinate AI across learning, teaching, research, LMS workflows, and administrative systems.
Students and faculty can encounter multiple AI tools. Policies, identity, models, and academic context often live in separate systems.
The challenge is the relationship between them.
ModelsProvider choices
StudentsLearning context
FacultyInstructional intent
LMSCourse workflows
PoliciesAcademic expectations
ITIdentity & access
02The recurring response
Reacting to AI can become an institutional operating model.
A reactive model can look like this: each new tool reopens the same questions about policy, procurement, access, and instructional use.
- 01A new model appears
- 02Students begin using it
- 03Faculty revisit policies
- 04IT evaluates another tool
- 05Leadership updates guidance
Another model appears. The cycle starts again.
When governance belongs to the tool, the institution stays one decision behind.
03An intentional alternative
An AI-native university operates differently.
It defines the conditions for AI use before the next tool arrives. Those conditions follow the institution’s structure, rather than being rebuilt inside every product.
AI-native means having the infrastructure to govern AI intentionally. People still decide where AI belongs—and where it does not.
Institutional intent becomes the starting point.
- 01Who?Identity & role
- 02Which AI?Models & providers
- 03Where?Institution, course, assignment
- 04For what purpose?Learning & operational intent
- 05Under what rules?Academic policy
04Institutional ownership
Own the rules, not necessarily the models.
Universities should be able to choose a provider, host a model, or consider a future option without starting their academic governance from scratch.
RedAPPL is designed so institutional governance does not have to be rebuilt around a single provider.
The provider supplies the model.
The university defines the conditions of use.
Underlying providers · examples
- OpenAI
- Anthropic
- Microsoft
- Local models
provider boundary
RedAPPLIdentity · Permissions · Academic context
University rules
applied in context
FacultyCourseAssignment
05From guidance to behavior
A policy students can ignore is only guidance.
Universities and instructors already set expectations. The missing layer is bringing those expectations into the AI environment itself. RedAPPL is designed to connect reviewed policy to the assistance a student can receive.
- Written policy
- RedAPPL
- AI behavior
Illustrative interaction
Student
“Write the introduction for me.”
RedAPPL
“This assignment allows brainstorming and outlining, but not draft generation. I can help you develop the argument and structure your introduction.”
06Context that carries through
AI should understand the structure of the university.
RedAPPL is designed around the structure universities already have. Permissions and policies can start at the institutional level and become more specific as the academic context narrows.
Inheritance means a course starts with the rules above it. An assignment can then define appropriate local choices within those boundaries, rather than starting with a blank policy.
Shared foundations. Context-specific decisions.
- 01
Institution
Approved models and baseline restrictions
A shared starting point - 02
Faculty / Department
Additional requirements
Within institutional boundaries - 03
Course
Course-level AI expectations
Connected to instructional intent - 04
Assignment
Specific allowed capabilities
Appropriate to the task
07Identity-aware access
Different roles shouldn’t receive the same AI experience.
A person’s role helps determine the models, capabilities, context, and data they should be able to access. These entitlements—the permissions that come with a role—are part of AI governance.
Professor
Teaching and course tools
A way to express instructional intent and configure the learning context.
Researcher
Institution-approved advanced AI access
Model and capability choices appropriate to the research context.
Admin / IT
Institutional configuration and oversight
A place to define access, model choices, and baseline governance.
08Continuity through change
Governance should outlive the model vendor.
Models will change. Providers, contracts, and capabilities will change with them.
University structure, academic expectations, identity, and permissions should not need to be rebuilt every time.
RedAPPL’s architecture separates institutional rules from the underlying AI provider.
- TodayModel A
- Next choiceModel B
- Future choiceModel C
University policy layer
Institutional intent remains.Structure · Identity · Permissions · Academic rules
09The control layer
What an AI-native university controls
Identity
Who receives access to which AI capabilities.
Models
Which providers and models the institution makes available.
Academic policy
How AI behaves inside courses and assignments.
RedAPPL control layer
Governed AI across the university
The university sets the terms
The future university won’t simply provide access to AI.
It will decide how AI works.
RedAPPL is the control layer designed to make that possible.
Explore Use Cases