What happened
Bar-Ilan University piloted an inside its Moodle learning platform across 71 courses and about 2,400 students. Built by KPMG and OpenApp with Microsoft Copilot Studio and Azure AI Services, it searches and summarizes staff-uploaded lectures, transcripts, presentations and articles, and can generate quizzes. The university plans to add it to another 100 courses during the first semester of the 2026/27 academic year. Price and access beyond participating courses are not documented.
Bar-Ilan University says it began piloting the agent in February on its Moodle platform. The system answers natural-language questions about course content, produces summaries and offers multiple-choice and open-ended quizzes. It is grounded exclusively in materials uploaded by teaching staff, including recorded lectures, transcripts, presentations and articles.
The pilot covered 71 courses and approximately 2,400 students. The agent is available in Hebrew, Arabic and English, and runs in the university’s Microsoft Azure environment. BIU says it plans to extend the system to another 100 courses in the first semester of the 2026/27 academic year, which began in October.
The source describes lecturers revising course materials, adding lecture transcripts and research papers, and testing responses before deployment. Lecturer Eliyahu Biton reported fewer routine questions and said the average final score in his class of roughly 50 students was three percentage points higher than in previous years, while acknowledging that the result could reflect a better-than-average cohort.
Source details: news.microsoft.com ↗
Why it matters
This is a concrete university deployment of an AI system positioned as a study aid and teaching assistant, rather than a general-purpose chatbot. The source says the agent is restricted to course materials uploaded by staff, operates in the university’s secure Azure environment and supports Hebrew, Arabic and English. That grounding may make it more useful for course-specific revision, but the evidence remains institution-reported: the source provides no independent accuracy assessment, controlled comparison or evidence that the tool caused improved grades. The reported three-point increase in one lecturer’s class may also reflect a stronger cohort.
The deployment gives students a course-specific way to retrieve information and practice answering questions, including material they may have missed. A student quoted by Microsoft said she reduced her reported study time after lectures from five or six hours to about three, though this is an individual account rather than an evaluated result.
For instructors, the source describes a shift toward updating and structuring teaching materials, monitoring answers and spending more time mentoring. That creates a practical implementation requirement: the agent’s usefulness depends on the completeness, clarity and maintenance of the materials supplied by staff.
The source’s evidence is not independent. It does not report a formal accuracy , a controlled study of learning outcomes, student participation rates, error rates or results across the three supported languages. The reported exam improvement is limited to one course and is explicitly confounded by cohort differences.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
crm_get_transaction(id='4092').An agent must create a draft calendar event for Tuesday at 2 p.m. Which evidence would establish the requested result?
What to watch next
The next useful evidence will be whether the expansion to 100 additional courses occurs as described, how staff maintain and audit the source materials, and whether students receive clear limits on the agent’s answers. Outcomes to watch include independent measures of accuracy, learning, workload and access across languages, as well as how the university handles privacy and academic-integrity questions. The source does not state whether the agent will eventually be available university-wide, whether students pay anything, or what safeguards apply to student data.
BIU’s planned expansion should clarify whether the system remains limited to participating courses or becomes broadly available to the university’s more than 20,000 students. The source does not document pricing, eligibility outside the pilot and planned courses, or any consumer access.
Because the agent can summarize and generate exams, independent evaluation should examine incorrect or incomplete answers, how or source passages are presented, and whether students can distinguish staff-approved material from generated explanations.
The source does not explain retention, access controls or other privacy arrangements for student interactions in the Azure environment. It also does not describe academic-integrity rules governing AI-generated study materials or assessments.