India uses AI to scan nearly 500 government apps for cyber vulnerabilities

India has begun using artificial intelligence to identify cybersecurity vulnerabilities across government digital infrastructure, with nearly 500 government applications and IT systems stress-tested over the past month, Electronics and IT Secretary S. Krishnan said.

Speaking to NDTV Profit, Krishnan said the exercise is being carried out by the Indian Computer Emergency Response Team (CERT-In), which has created a dedicated AI sandbox to test government systems against potential cyber threats.

"We've been at it for about a month. CERT-In has created a sandbox where multiple AI models are being used to test government applications and systems," he said.

According to Krishnan, the government is using around nine or 10 leading open-source AI models along with one or two Indian-developed models to identify security weaknesses before they can be exploited.

The AI models are being used to probe applications for vulnerabilities, assess system resilience and detect security flaws that conventional audits may overlook.

"So far, close to 500 government applications and systems have been stress-tested, and the work is continuing," he said.

The initiative is part of the government's broader effort to integrate AI into public administration while strengthening cybersecurity as India's digital public infrastructure continues to expand.

The move comes as governments worldwide increasingly deploy AI for cybersecurity, using the technology to automate vulnerability detection, threat analysis and incident response even as generative AI enhances the capabilities of cyber attackers.

India has also been expanding AI infrastructure under the IndiaAI Mission while working on regulatory frameworks for the responsible use of the technology.

Krishnan said the use of AI to secure government systems demonstrates how the technology can be deployed in the public interest as policymakers continue to deliberate on broader AI governance and safety frameworks.