US Government Wants to Check AI Models

May 5, 2026, 17:48

Introduction

The US government wants to check new AI models before people use them.

Main Body

A new AI model called Mythos can find problems in computers. The government is afraid that bad people will use this AI to attack. The White House wants to make a group of experts to test these models first. Some companies like Google and Microsoft agree to this. But a company called Anthropic does not agree. The government and Anthropic are now angry and do not speak well. Workers at Google DeepMind in the UK are also unhappy. They do not want AI to help the military. They want to start a union to stop AI from making weapons.

Conclusion

The US government wants more control over AI, but many workers disagree.

Learning

💡 The 'Feeling' Words

In this story, people have different emotions. Look at how we describe them:

Afraid $\rightarrow$ Scared of something bad.
Angry $\rightarrow$ Very mad.
Unhappy $\rightarrow$ Not happy/Sad.

🛠️ How to say 'No'

To reach A2, you need to show disagreement. Notice these two ways from the text:

Direct: "...does not agree."
Emotional: "...are now angry."

Tip: Use do not or does not to flip a sentence from 'Yes' to 'No'.

Example: They want AI $\rightarrow$ They do not want AI.