India seeks access to Anthropic's Mythos while using other AIs
India is trying to get its hands on Mythos, a powerful AI from Anthropic that can spot hidden bugs in popular software, stuff that usually slips through the cracks.
While talks with the US and private companies are ongoing, Indian teams are using other AI models (with about 60% to 70% of Mythos's skills) to keep things secure for now.
As S Krishnan from the IT Ministry put it, this tech could really help fix issues before they become big problems.
IT Ministry creates AI testing sandbox
To prep for advanced AIs like Mythos, the government has created a special sandbox environment where they can safely test out these models on real code.
Krishnan also made it clear that any super-sensitive data will stay within India's own servers, not on foreign clouds, to keep critical systems safe while waiting for access to next-generation technology.