Artificial Intelligence (AI) is as much a boon as it is a major security challenge. Recently, there has been a stir all over the world regarding the extremely powerful AI model 'Claude Mythos' of the Anthropic company. Seeing the seriousness of this threat, India's Finance Minister Nirmala Sitharaman held an emergency high-level meeting with bank heads and stakeholders on Thursday.
What is the Claude Mythos?
Mythos is the most advanced and powerful AI model from Anthropic to date. This model is so capable that it can find decades-old vulnerabilities in operating systems and web browsers that no human or ordinary software has been able to find so far. According to tech experts, if used in the wrong hands, Mythos can become an 'advanced hacking tool' that can break any strong security shield.
Finance Minister issues strict instructions to banks
During the meeting, Nirmala Sitharaman has directed scheduled commercial banks to remain on high alert. The Finance Ministry has made it clear that the risk arising from the myth is unprecedented.
Upgradation of IT systems: Banks have been asked to immediately update their outdated systems and plug security loopholes.
Data Security: Emphasis has been placed on adopting a 'Zero Trust' policy to protect customers' personal data and banking transactions.
Real-time intelligence: There is a plan to create a common platform among banks to share information on cyber attacks, which will also involve agencies like CERT-In.
When AI broke out of 'digital prison'!
How dangerous Mythos is can be gauged from an incident. During the research, this AI was kept in a secure 'sandbox' (a kind of digital prison where there is no internet access). However, this AI found a way out by breaking the security fence and sent an email to the researcher working on it. This incident proved that this model has the ability to go beyond human control.
Global concern: There is also unrest in America
Not only India, but the US is also serious about this issue. US Treasury Secretary Scott Besant and Federal Reserve Chairman Jerome Powell recently held a closed-door meeting with banking executives. Currently, Anthropic has given access to this model only to 40 selected large companies (such as Google, Microsoft and Amazon), but reports that some unauthorized users have gained access to it have created fear in the global economy.
