World News for Students of English

26-05-2026 15:00

A company named Anthropic made a new AI model called Claude Mythos.

This AI is very good at computer code and cybersecurity. It can find mistakes in computer systems very fast. It can even combine four or five small mistakes to make a big cyberattack.

This is dangerous because hackers could use it to attack power companies, stock exchanges, or the military. Because Claude Mythos is so powerful, Anthropic doesn’t give it to everyone. Only a small group of partners in the US can use it to fix security problems.

Other countries, like India, are not happy. They worry that without this AI, they can’t protect their computers. In February 2026, world leaders met to talk about global rules for AI, but it is very hard for countries to agree, because they want to keep their technologies safe.

Difficult words: cybersecurity (protecting computers, servers, and networks from digital attacks), stock exchange (a place where people buy and sell parts of companies), security (being free from danger or threat).

You can watch the original video in the Level 3 section.