The Indian Computer Emergency Response Team (Cert-In) has operationalised an AI war room to test advanced AI models, identify vulnerabilities, and prepare India for wider access to frontier artificial intelligence technologies, according to IT Secretary S Krishnan. Speaking during the launch of the second edition of Cert-In's Digital Threat Report, Krishnan explained that the war room operates with AI models that have 60-70% capability of frontier AI models such as Anthropic's Claude Mythos or OpenAI's ChatGPT 5.6, among others. The initiative represents a significant operational advancement from India's previous sandbox development efforts, with Cert-In creating sandboxes and testing environments where they are testing various elements with these AI models and identifying possible vulnerabilities.
While pursuing access to Mythos and other advanced frontier models, India is simultaneously using alternative AI models in a CERT-In sandbox to detect software vulnerabilities and strengthen cybersecurity, as reported by Rediff Moneynews. IT Secretary S Krishnan explained that these alternative models are being leveraged to track and fix vulnerabilities while negotiations with US counterparts continue. The sandbox environment is designed to test code, identify vulnerabilities, and develop secure workflows before access to more powerful models becomes available. According to Krishnan, "The entire exercise is akin to a dry run, so that we, in a sense, have done a dry run for whenever Mythos is available in order to do it in a secure environment, and at the same time making sure that at least a significant proportion of vulnerabilities have been identified and tackled even without access to Mythos."
According to Rediff Moneynews, Mythos - an advanced AI model developed by Anthropic - has been able to "lay bare" vulnerabilities that existed in the original code of widely-used software programmes but were often left unresolved. Krishnan noted that these vulnerabilities were considered low priority, too costly to fix, or had simply gone undetected. The government's approach focuses on ensuring these models are accessible within India's borders, addressing the constraints imposed by international export restrictions. As Krishnan explained, "What this gives is an opportunity now to actually identify many of these vulnerabilities and correct them."
The war room will also ensure that a significant proportion of the development of indigenous AI models and protection mechanisms is done without relying heavily on these frontier general-purpose models, according to Krishnan. This strategic approach addresses the government's concerns about dependency on foreign AI technologies. As Krishnan noted, "There is a risk, of course. The government is concerned that we should have access to some of these tools, not on a cloud-based approach and sending it out...At least some most sensitive elements of code will have to be done 'on-premises' within the country." In India, access to Anthropic's latest general-purpose model Claude Mythos was initially extended to a few government agencies such as the Indian Cyber Crime Coordination Centre (I4C), Cert-In, the National Critical Information Infrastructure Protection Centre, and the Digital Intelligence Platform of the Department of Telecommunications as part of the expansion of Project Glasswing.
Access to Anthropic's Claude Mythos was later halted after the US imposed restrictions on foreign persons' access to the model, as reported by Business Standard. The government is currently in talks with US counterparts and respective companies to restore India's access to these models. Krishnan emphasized that "Getting access to Mythos and similar advanced frontier models is very high on the priority of the government, and this is something that we have taken up with our counterparts in the US and with the respective companies." He noted that there is a process involved in the negotiations, stating "We are negotiating that process; we are going through that right now."
