← Cases
United StatesEducation2024 – ongoing

Syracuse University

Campuswide Claude access and AI-assisted learning exercises

A Syracuse instructor reported exam scores 12 points above previous results after redesigning Claude practice exercises

WorkingVendor case studyVendor ↗

Problem

An instructor struggled to give students individual feedback in a large class. An initial Claude-generated multiple-choice practice exam produced little improvement in scores.

Approach

Syracuse provided Claude access across the university over a 2-year rollout. After consulting education specialists, the instructor replaced multiple-choice practice with written definitions. Claude assessed responses against lecture material and returned feedback. The university also launched Clementine, a Claude-powered class-search tool connected to institutional data.

Results

12 pointsexam-score increase above previous results reported by the instructor in Anthropic's interviewVendor case study[1]

Takeaway

The reported score increase followed a change in practice format. The university did not establish how much of the improvement came from Claude itself.

Categories

Industry
Education
Innovation type
Process innovation
AI technology
Generative AI
Evidence
Vendor case study

Sources

Source summaries

[1]Anthropic1 Apr 2026

Anthropic interviewed Syracuse's chief digital officer about campuswide Claude access and teaching applications. He reported a 12-point exam-score increase after changing practice exercises and described the launch of a class-search tool.

Full details are in the original sources.

Added 1 Nov 2026 · Updated 1 Nov 2026
Open case page ↗