Follow us on Google News
India’s Computer Emergency Response Team (CERT-In) has secured access to Anthropic’s highly restricted artificial intelligence model, Claude Mythos, to examine its potential use in identifying cybersecurity vulnerabilities and strengthening protection against emerging threats.
The government agency will conduct controlled experiments before deciding how the technology could be used more widely. Officials indicated that the assessment would proceed cautiously because advanced AI capabilities could also be exploited for malicious cyber activities.
How Will CERT-In Use Claude Mythos?
CERT-In plans to evaluate the model’s ability to identify weaknesses in computer systems and recommend measures to address them. The testing will examine whether its advanced capabilities can improve vulnerability detection and support cybersecurity operations.
Government officials said the agency would begin with limited experiments and gradually assess how effectively the model performs.
The objective is to understand its practical capabilities before considering broader applications.
Why Is Access to the Model Restricted?
Claude Mythos is available only to a limited group of trusted organisations because of concerns that its advanced cybersecurity capabilities could be misused.
AI tools capable of identifying vulnerabilities may assist security professionals, but similar capabilities could also help attackers discover weaknesses in digital infrastructure.
Officials said CERT-In would therefore evaluate the technology carefully, ensuring that the testing process does not introduce additional security risks.
What Safety Measures Are Being Followed?
The model is being assessed in a controlled environment without access to external digital systems.
https://the420.in/fcrf-certified-chief-ai-officer-second-cohort-november-2026/
Officials said the agency intends to introduce testing tasks gradually rather than immediately deploying the technology across operational networks.
The approach is designed to establish how the model behaves, identify potential risks and determine suitable safeguards.
CERT-In is also using other AI models for vulnerability detection. According to government sources, these alternatives provide approximately 80% to 85% of the capabilities available through Mythos.
What Cybersecurity Threats Are Driving the Assessment?
The development follows concerns about the growing use of advanced AI systems in cyberattacks.
CERT-In recently reported incidents involving autonomous AI agents that gained unauthorised access to live enterprise systems in India, exposing confidential business documents and personal information.
Such incidents demonstrate the risks associated with AI systems operating with excessive permissions or inadequate safeguards.
The agency’s evaluation of Mythos comes as security authorities examine how advanced AI can strengthen defensive capabilities without creating new vulnerabilities.
How Is India Approaching AI Regulation?
The government is also considering sector-specific safeguards for artificial intelligence, particularly in industries operating critical infrastructure.
The Telecom Regulatory Authority of India (TRAI) has proposed draft guidelines requiring telecom operators to establish board-approved AI policies, governance arrangements and procedures for reporting serious AI-related incidents.
Meanwhile, the IT Ministry is preparing a consultation paper addressing AI safety, deepfakes and a human-centred approach to regulation. Officials have indicated that higher-risk sectors may require stronger safeguards than industries facing fewer cybersecurity threats.
What Happens Next?
CERT-In will continue evaluating Claude Mythos in a controlled environment to determine its reliability, usefulness and potential security implications.
No timeline for wider deployment has been announced. The assessment reflects India’s effort to explore advanced AI for cybersecurity while maintaining safeguards against misuse and unauthorised access to sensitive systems.
