--- title: "Security-First LMS Integration" slug: "security-first-lms-integration" author: "Jeremy Weaver" date: "2025-08-21 16:37:21.743285" category: "Premium" topics: "LTI 1.3 integration LMS AI integration Canvas LTI Developer Key Blackboard Ultra LTI tool Brightspace LTI Advantage OpenID Connect LTI launch JWK URL keyset Target Link URI External Tool Activity Assignments and Grade Services (AGS) Names and Roles Provisioning (NRPS) Course-scoped AI agent Retrieval-augmented generation (RAG) LMS API ingestion Instructor-controlled AI Secure AI in higher education Privacy-first AI tutoring AI gradebook integration LTI placements and deployments University AI governance" summary: "A practical, standards-aligned overview of how ibl.ai integrates with Canvas, Blackboard, and Brightspace using admin-registered LTI 1.3, optional, IT-approved RAG ingest, and course-scoped links—delivering security, transparency, and instructor control without fragile workarounds." banner: "" thumbnail: "" --- # Security-First LMS Integration — LTI 1.3, Explicit RAG, and Course-Scoped Control Campus IT teams don’t need another “magic” integration; they need one that’s predictable, auditable, and easy to govern. That’s why ibl.ai connects to your LMS the standards-aligned way: **LTI 1.3 / LTI Advantage** for the launch, **admin-approved API keys** (only if you want content ingestion), and **course-scoped links** so access is exactly as wide—or narrow—as you intend. No scraping, no shadow permissions, no surprises. --- ## Start With The Standard, Not a Workaround ibl.ai is registered as an **LTI 1.3** tool at the institution level. Your LMS admin adds our OIDC login, redirect (target link) URIs, and keyset (JWK) URL, then chooses placements the same way they would for any other trusted tool. That gives you a well-understood security posture (signed launches, role/context claims), plus clear toggles for what’s on and what stays off—like **Assignments & Grade Services (AGS)** and **Names & Roles (NRPS)**. **The result**: instructors click a link, we get the sanctioned LTI launch, and the agent opens with the right context—nothing more. ## What This Looks Like In Each LMS - **Canvas**: Admins create a Developer Key (LTI 1.3) and add the app by Client ID. Instructors place the agent as a **Module → External Tool** item or in course navigation. - **Blackboard (Ultra & Original)**: In Ultra, instructors add **Teaching Tools with LTI Connection**; in Original, **Build Content → Web Link** marked as LTI. Visibility controls keep student access explicit. - **Brightspace (D2L)**: Admins **Register Tool** under LTI Advantage, then **Create a Deployment** and **Create a Link** to the specific agent. Instructors add it as an **External Learning Tool** activity. Across all three, the pattern is the same: the **link points to a specific agent**, which keeps scope tight and predictable. ## Retrieval-Augmented Generation (RAG) the Explicit Way We support two secure, opt-in paths for grounding answers in authoritative materials: - **Instructor-managed uploads**. Faculty drag-and-drop approved course files (syllabi, slides, PDFs) into ibl.ai. RAG cites those exact sources back to students. No LMS API access is involved. **IT-approved LMS ingestion**. If you want agents to index course content directly from the LMS, your IT team provisions scoped LMS API credentials. Only then do agents pull permitted files. Nothing is auto-ingested by default. Either way, every response is **citation-first**, so instructors and students see where information came from. ## Course-Scoped Access and Role-Aware Behavior A single LTI link maps to **one agent**. That agent’s knowledge base, safety policies, and features are defined at the course (or department) level. LTI roles determine what users can do—e.g., instructors see analytics and settings; learners just chat and complete activities. If you enable AGS, agents can post scores back to the LMS; if you leave it off, agents remain read-only. ## Why Security Teams Like This Model - **Familiar controls**. You manage keys, placements, and scopes using the tools your LMS already provides. - **No hidden data flows**. We don’t crawl the LMS; ingestion only happens with **explicit uploads** or **admin-approved API credentials**. - **Easy to roll out, easy to roll back**. Start in a few courses, add links as confidence grows, and only enable gradebook or roster access when there’s a clear need. - **Clear auditability**. LTI launches and any optional API calls are visible to your LMS admins. Within ibl.ai, every grounded answer carries source citations. ## A Simple Path to Pilot (and Scale) - **Pick one course** where an agent adds immediate value (e.g., large enrollment or heavy Q&A). - **Register ibl.ai via LTI 1.3** and add the course link. - **Choose your RAG path**: instructor uploads only, or IT-approved LMS API ingestion. - **Run for the duration of the pilot**. Instructors keep full control (factory defaults that just work; deeper prompt/pedagogy settings when they want them). - **Review analytics and citations** with faculty and IT, then decide where to expand next. --- ## Want To See This in Action? - **Short walkthroughs**: Our setup guides for Canvas, Blackboard, and Brightspace show the exact admin steps and placements we support, and can be found at **[ibl.ai/docs](https://ibl.ai/docs/os/overview)**. - **Live demo**: We’ll embed an agent in a sandbox course and ground it on your materials so you can test citation quality and controls. - **Next steps**: DM us on LinkedIn or visit **ibl.ai/contact** to schedule a 20-minute fit check with your LMS admin and security lead. ibl.ai integrates the way your LMS expects, gives instructors real control, and keeps your governance clean—so you can move fast without cutting corners. ## Why does owning the AI stack matter? **ibl.ai is the agentic AI platform where you own all the code and the data. You self-host the entire stack inside your own perimeter, run it model-agnostic across any LLM and switch anytime, and pay by usage with no per-seat pricing — so you can deploy anywhere: your cloud, on-premise, GovCloud, or fully air-gapped.** - **You own all the code and the data.** Full source code under a perpetual license, running on your infrastructure. Not API access to someone else's platform — the stack itself is yours. - **Model-agnostic.** Run any LLM — Claude, GPT, Gemini, Llama, Command, or your own fine-tune — and switch providers without rewriting the platform. - **No per-seat pricing.** Usage-based billing against a budget cap you set. Cost tracks what your organization actually uses, not how many people you employ. - **Deploy anywhere.** Your cloud, your VPC, on-premise, GovCloud, or a fully air-gapped network with no outbound connectivity. 1.6M+ users across 400+ organizations run the platform this way, including NVIDIA, MIT, and Syracuse University. ibl.ai is family-owned and operated from New York, NY — a U.S.-headquartered, domestically-owned long-term partner, not a vendor that sells licenses and moves on.