Govt In Talks With US For Access To Anthropic Mythos AI Amid Security Concerns

The Government of India is in discussion with the administration of the US and the American tech company, Anthropic, to seek access to the Mythos AI model. These talks are underway as India aims to ensure fair access for the Indian companies and to safeguard the critical infrastructure, as per The E

What is the story about?

govt in talks with us for access to anthropic mythos ai amid security concerns

The

Government of India is in discussion with the administration of the US and the American tech company, Anthropic, to seek access to the Mythos AI model. These talks are underway as India aims to ensure fair access for the Indian companies and to safeguard the critical infrastructure, as per The Economic Times (ET).

Govt Wants To Seek Fair Access To Mythos AI

A government official told ET that the center wants access to Mythos to be 'equitable' so that it can ensure a level playing field. Moreover, the government is also focusing on safeguarding key systems such as power grids, banking channels, and telecom networks that could be exposed to the new emerging AI-led risks. Speaking at the ET Awards for Corporate Excellence, the union finance minister Nirmala Sitharaman stated the Ministry of Electronics and IT is constantly engaging with US officials and vendors and testing the model. She warned that the cyber risk associated with Mythos AI could be significant. The FM recently held a meeting with the IT minister, Ashwini Vaishnaw, and banking officials along with cybersecurity regulatory bodies to review preparedness. Apart from the US officials, the government is also reportedly engaging with Anthropic at its US headquarters. Agencies such as the Indian Computer Emergency Response Team and the National Critical Information Infrastructure Protection Center have been told to protect the critical infrastructure that could potentially be impacted amidst the evolving risk from emerging AI models.

Why Mythos AI Could Be Dangerous

Anthropic has restricted the use of its Mythos to select groups of companies. Under its Project Glasswing, nearly 40 US companies were given access to this model, as it is seen as a highly advanced AI system that is capable of identifying and exploiting critical software vulnerabilities. In simple words, this AI tool can not only identify software vulnerabilities but can also be used as a hacking mechanism on its own to access the device, especially banking software, noted Anthropic. Reportedly, Nasscom has written to Anthropic to seek access for Indian companies to help safeguard the cyber resilience. Additionally, officials are also considering wider policy responses, as more tech players could launch similar AI models in the future.