NAAC to Introduce AI-Based Accreditation System, Discontinue Physical Inspections

The National Assessment and Accreditation Council (NAAC) is set to introduce an AI-powered system starting in August. The reform eliminates traditional peer visits and physical inspections, replacing them with a fully digital, machine-learning-based model under a new Binary Accreditation system.

NAAC executive committee chairman, Anil Sahasrabudhe, reportedly said that the AI-backed approach aims to enhance transparency, reduce delays and significantly expand the number of accredited institutions. “The new framework eliminates peer team visits, which often led to delays and integrity concerns. AI and stakeholder feedback will now drive credibility,” he said.

Institutions that qualify can then opt for a second tier called Maturity-Based Graded Accreditation, which includes five levels. From Level 3 onwards, physical visits will resume in a hybrid format to reduce manipulation risks.

A key innovation is an AI-powered credibility score. Institutions begin with a default score of 0.5, which is adjusted based on the validation of submitted documents. The AI system cross-verifies data, and a panel of over 100 stakeholders, which includes faculty, industry experts, retired VCs and NGOs, assesses authenticity. Fake submissions can result in a three-year accreditation ban.

This shift is based on recommendations by a committee led by former ISRO chief Koppillil Radhakrishnan, formed in 2022 to revamp India’s accreditation ecosystem. Currently, only 40% of universities and 18% of colleges are accredited. The new system aims to bring over 90% of higher education institutions (HEIs) under accreditation in the next five years.

Foreign universities planning to set up campuses in India will also be eligible under this new framework.

NAAC officials reportedly said the AI-led model will ensure consistency and scalability. However, some educationists have raised concerns about whether AI alone can capture the realities of different institutions.