But AIs can still be tricked.

For instance, a couple of months ago, researchers at Palo Alto Research, they discovered the simple way to break through the guardrails was to use terrible grammar and no punctuation and have a sentence which has no full stop at the end, which would be the normal point at which the guardrail would have a chance to kick in before the jailbreak takes effect

and so just as long as you keep on and on on on nona nona nonna nona nonna no no no no no no no please interrupt me mark

I've just thinking I've worked with programmers who use exactly the same tactic

smashing security Smashing Security, Episode 433,

How Hackers turned AI into their new henchmen with Graham Cluley.

Hello, hello, and welcome to Smashing Security episode 433.

My name's Graham Cluelly.

And I'm Mark Stockley.

Mark, lovely to have you on the show again. Of course, listeners will know about you. But for those who haven't yet experienced you, you are not only the co-host on that marvellous podcast known as

the AI Fix. That's true. Alongside me, but you're also the cybersecurity

evangelist at Threatdown. That's true too. And for people who haven't heard of Threat Down,

what is that? So Threat Down is the business product side of Malwarebytes. So probably everybody's

heard of Malwarebytes. Yes. Lots of people have got great stories about using

malware bites to battle virus infections back in the day. And Threatdown is basically that

stuff but then built for businesses. Oh, cool. Well, before we kick off, let's thank this

week's wonderful sponsor Vanta. We'll be hearing more about them later on the podcast.

This week on Smashing Security.

We won't be talking about how British luxury carmaker Jaguar Land Rover has been hit

by a cyber attack, thus severely affected its retail and production systems.

You'll hear no discussion of how the US Department of Homeland Security has fired 24 people

after hackers broke into FEMA, the Federal Emergency Management Agency,

...