AI Assessment Tool Piloted Across 20 Indian Schools Shows Promise for Underserved Students

Last updated: February 27, 2026

If you’ve ever watched your child struggle with practice exams, knowing they needed more feedback than any overstretched teacher could provide, you’re witnessing exactly the problem a new partnership is designed to solve. Anthropic and Pratham Education Foundation have deployed a Claude-powered assessment tool across 20 Indian schools, and early results suggest this approach could transform how we support learners who lack access to individualized instruction. This isn’t about replacing teachers—it’s about giving them superpowers.

TL;DR

Anthropic partnered with Pratham Education Foundation to deploy a Claude-powered AI assessment tool across 20 Indian schools.
The Anytime Testing Machine serves 1,500 students currently, with plans to expand to 100 schools by end of 2026.
Grading accuracy improved from 30% to 80% through iterative prompt engineering; question generation achieves 90% accuracy.
Teachers remain final evaluators, reviewing AI-generated feedback before it reaches students.

AI Assessment Tool Reaches 1,500 Students in India

Anthropic has partnered with Pratham Education Foundation—one of India’s largest education nonprofits—to deploy a new AI-powered formative assessment tool called the Anytime Testing Machine (ATM). The system, powered by Anthropic’s Claude model, generates curriculum-aligned questions, grades handwritten responses, and delivers personalized feedback to students.

The pilot program is currently running with 1,500 students across 20 schools, with plans to expand to 100 schools by the end of 2026. The tool has also been adapted for Pratham’s Second Chance program, which supports women preparing for India’s grade 10 board exam and currently serves more than 5,000 learners.

A new AI tool helped 1,500 Indian students get personalized feedback on their work—turning one teacher into many. The results are promising. Tweet Now Follow Learning Success

Grading Accuracy Improved from 30% to 80%

Pratham spent three decades working to close learning gaps across India, but one persistent challenge was grading. In many classrooms, teachers manage 60 or more students, limiting their ability to provide individualized feedback on practice exams. The challenge was especially acute in the Second Chance program, where women who left formal schooling lack access to trained subject-matter instructors.

Through iterative prompt engineering and evaluation design, the grading accuracy improved from approximately 30% to roughly 80% alignment with subject-matter experts. On question generation, the system now delivers 90% accuracy aligned to Bloom’s Taxonomy. The Claude-powered ATM has completed more than 1,500 student assessments, and the Second Chance program plans to migrate fully to the Claude-powered pipeline by the end of 2026.

Author Quote "

Quote: Teachers feel empowered because they remain the final evaluators who validate AI feedback before it reaches the student. Attribution: Sravana Chandra, AI Lead at Pratham

Not applicable - no significant bias identified

Teachers Remain Central to the Assessment Process

A core principle of the design is that teachers review and can refine AI-generated feedback before it reaches students. Sravana Chandra, AI Lead at Pratham, notes: “Teachers feel empowered because they remain the final evaluators who validate AI feedback before it reaches the student.”

Multilingual capability has also been central to deployment. Given Claude’s linguistic capabilities, it was readily able to handle feedback generation in a mix of Hindi and English, using English terms where relevant (such as for scientific terms) while keeping most of the text in Hindi. Anthropic and Pratham are now expanding the partnership to support Tech in TaRL (Teaching at the Right Level), with a randomized controlled trial planned for several thousand students.

Key Takeaways:

AI Assessment Pilot Expands: Claude-powered tool deployed across 20 Indian schools with 1,500 students, targeting exam preparation gaps in underserved communities.

Accuracy Improves Dramatically: Grading accuracy rose from 30% to 80% through iterative prompt engineering, with question generation achieving 90% Bloom's Taxonomy alignment.

Global Expansion Ahead: Partnership plans expansion to 100 schools by end of 2026, with potential rollout to Kenya, Rwanda, and other Global South regions.

Partnership Signals Broader Potential for Global South

The collaboration aligns with Anthropic’s broader India strategy. The company recently opened a Bengaluru office and announced partnerships across enterprise, education, and agriculture. India is now the second-largest market for Claude.ai, with educational and instructional tasks accounting for 12% of usage.

The organizations are also exploring digital public infrastructure initiatives, including knowledge graphs, and potential expansion to Kenya, Rwanda, and other regions in the Global South. This represents a significant step forward in using AI to address educational inequality—not by replacing human teachers, but by amplifying their capacity to provide personalized support at scale.

Author Quote "

Quote: Human agency is something we focus on. Instead of asking what AI can do without teachers, we ask: how can AI help our teachers and students exactly where they’re stuck? Attribution: Nishant Baghel, Director of Technology Innovations at Pratham

Here’s what matters most: this technology demonstrates that AI can be a powerful tool for expanding educational opportunity rather than replacing human connection. When teachers are empowered with AI that handles grading at scale, they can focus on what they do best—building relationships, inspiring curiosity, and providing the human guidance that no algorithm can replicate.

The research is clear: the most effective learning happens when children receive timely, specific feedback on their work. For too long, that level of personalization was only available to students whose families could afford private tutoring. AI has the potential to democratize that advantage—while keeping teachers at the center of the learning process.

If you’re ready to explore how Learning Success can help your child receive the personalized support they deserve, our free trial includes a personalized Action Plan—and you keep that plan even if you decide it’s not the right fit.