Marshall University adopts sovereign AI platform from ibl.ai

Marshall University is deploying ibl.ai’s Agentic AI Operating System on university-controlled infrastructure, giving the school full source-code and data ownership. The move is meant to help Marshall test AI across student services, measure ROI, and keep sensitive institutional data on campus.

Why it matters: - Marshall University is moving AI inside its own infrastructure instead of renting a vendor-hosted service. - The setup gives Marshall full control over code, data, access, security and model selection. - The university can evaluate AI use cases while keeping sensitive institutional data on campus.

What happened: - Marshall University announced the deployment from the main stage of the EDUCAUSE 2026 Annual Conference in Denver. - Marshall Chief Information Officer Dr. Jodie Penrod said the project fits into the university’s strategic planning around AI in teaching, learning and workforce preparation. - The university is using ibl.ai as a sandbox for faculty, staff and students to test AI tools and measure return on investment.

The details: - The agreement transfers the full ibl.ai codebase to Marshall, including the agent framework, orchestration layer, APIs, supporting services, LMS integrations through LTI, documentation and deployment guides. - Marshall’s IT team governs access, security and model choice because the university owns both the code and the data. - No outside party trains on Marshall’s institutional data. - The initial rollout includes higher-education AI agents for admissions, transcript evaluation, financial aid guidance, academic advising, tutoring and IT support. - Marshall will review those agents and prioritize the ones that best match institutional needs. - ibl.ai engineers will work with Marshall staff to integrate selected agents with university systems and build an AI automation layer across departments. - Marshall staff and developers also receive the environment needed to build department-specific tools, workflows and automations on top of the platform. - The agreement includes onboarding and hands-on training so the capability stays with Marshall. - The platform logs every interaction in Marshall’s own environment, letting the university track usage, users and cost. - Marshall can compare language models from different providers on its own data. - The platform is model-agnostic and priced on usage rather than per seat, so Marshall pays for actual consumption and can switch models as pricing changes. - ibl.ai is in production with more than 1.6 million users across more than 400 organizations. - The platform is SOC 2 Type 2 certified and can be deployed to private cloud, on-premise and air-gapped environments.

Between the lines: - Penrod’s comments point to a broader campus AI strategy built around control, measurement and internal capability-building rather than one-time software adoption. - The ownership model lowers vendor lock-in because Marshall keeps the code even if it changes providers later. - ibl.ai is positioning sovereign AI as a response to university concerns about cost, data privacy and long-term flexibility. - Marshall joins Syracuse University and other institutions using ibl.ai as owned infrastructure rather than rented software.

What's next: - Marshall will test and prioritize the first wave of agents before expanding use across departments. - The university and ibl.ai will work together on system integrations and workflow automation. - Marshall will use the platform to measure which AI tools deliver value for students and staff and to guide future adoption decisions. - The university can expand or change models as pricing and needs evolve.

The bottom line: - Marshall is betting that owning the AI stack will be cheaper, safer and more adaptable than licensing a conventional campus AI tool.