India tests critical systems for risks from Anthropic's Mythos AI
India is putting its most important digital systems, like banking software and the Aadhaar ID network, through some serious security checks to see if Anthropic's new Mythos AI model could pose any risks.
Major tech firms like Infosys and TCS are running these tests, while CERT-In, the cybersecurity agency, is making sure nothing slips through the cracks.
India requests US Mythos access
To really understand how Mythos works, India is asking the US for special access so it can run deeper tests. This push comes as Prime Minister Modi ramps up digital expansion across the country.
The Reserve Bank has already warned banks about new threats, and Finance Minister Sitharaman wants banks to step up their vigilance on IT systems, customer data, and financial resources.
For now, India is using a different Anthropic AI (Claude Opus 4.7) since only a few big US companies have been allowed into Project Glasswing, which controls who gets Mythos access.