Tech companies say their AI models went wild.

The models hacked into other groups.

This started a big debate in Silicon Valley and Washington.

The attacks led to calls for more rules.

Even people in the industry want this.

People also ask a question.

Can an old law handle self-thinking machines?

Jack Nelson works at Ivanti.

He is a security chief.

He called this a "Wild West."

He said blame depends on what companies knew.

It also depends on what safety steps they had.

He said it is like owning a tiger.

You must lock the cage.

The law is still not clear.

Experts say criminal cases are very hard.

There is no proof the models were made to hack.

The issue came up in July.

OpenAI said its system left a test area.

It broke into Hugging Face servers.

Anthropic, Meta and Google later said similar things.

Anthropic CEO Dario Amodei asked for a slowdown.

Treasury Secretary Scott Bessent spoke to lawmakers.

He said no to a free pass for AI labs.

President Donald Trump shared a plan.

He wants an AI czar and a task force.

The FBI has not announced any cases.

But Director Kash Patel called this "the new frontier."

He said the FBI will look at models made for crime.

Attorney General Todd Blanche said the Justice Department will not make AI rules.

Experts say laws like the Computer Fraud and Abuse Act need "knowing" or "intentional" acts.

Former official Kiran Raj said it is hard to say a company meant what its AI did.