AI agents being tested by OpenAI involved in cyber-attack on another service, say researchers

60 points · 10 comments · view on lemmy.world

10 Comments

gens@programming.dev · 30 pts · 5d (1 reply)

When prison?

Never. This is the US.

FauxLiving@lemmy.world · 29 pts · 5d (2 replies)

I'm pretty sure this is a crime, you can't 'oopsie, it wasn't me it was my computer' your way out of a cybercrime.

you can’t ‘oopsie, it wasn’t me it was my computer’ your way out of a cybercrime.

You can't, but they can.

There are those whom the law binds but does not protect and those whom the law protects but does not bind, as the saying goes.

They rich. They oppsie.

This is to get the military funding?

An agent can’t tell itself to go commit crimes

FIND ME A THROAT TO CHOKE

This is pro-OpenAI propaganda

The gall they have to publicly brag about their crimes.

I'll take "Even more bullshit" for 800, Alex