A hacker used AI to automate an 'unprecedented' cybercrime spree, Anthropic says

NBC News ^ | 8/27/25 | Kevin Colliar

[CaptainK]

A hacker has exploited a leading artificial intelligence chatbot to conduct the most comprehensive and lucrative AI cybercriminal operation known to date, using it to do everything from find targets to write ransom notes.

In a report published Tuesday, Anthropic, the company behind the popular Claude chatbot, said that an unnamed hacker “used AI to what we believe is an unprecedented degree” to research, hack and extort at least 17 companies.

[CaptainK]

Well this isn't good

[CaptainK]

Anthropic declined to name any of the 17 companies, but said they included a defense contractor, a financial institution and multiple health care providers. The stolen data included Social Security numbers, bank details and patients’ sensitive medical information. The hacker also took files related to sensitive defense information regulated by the U.S. State Department, known as International Traffic in Arms Regulations.

[ConservativeMind]

This takes “Script Kiddies” to a whole new level.

[Big Red Badger]

Who’d a Thunk

[lightman]

Time for our B-2s to hit some “data centers”.

[CaptainK]

If the smartest people ever are developing and running these AI chatbots what do you call the people who are able to thwart their threat security systems?

[lightman]

Smarterasses?

[mass55th]

"Well this isn't good"

Any technology like this created by humans, can and will be manipulated by other humans and used to commit crimes.

[Callnote]

Thats Eagle eye

[CaptainK]

I’d copyright that if I were you.

[rarestia]

As someone who dabbled in “script kiddie” circles as a youth, today’s youth are learning how to use AI to their advantage. While we sat in dark bedrooms poring over batch files, they are “talking” to AI to find where organizations hastily implemented something to take advantage of “lazy admin” practices and walk right in the front door with stolen credentials.

Make no mistake, AI isn’t necessarily doing the heavy lifting, but like an overly-friendly golden retriever, it has no problem showing the burglars where to find the expensive jewelry.

[exinnj]

Any advanced technology developed by humans will be first used to promote and exploit pornography and then to commit crimes against other humans.

[The Duke]

The perps could just as easily have visited a web page with a ‘howto’ list. AI is just incidental here.

[yelostar]

We’ll be seeing a lot more of this..

[yelostar]

Gematria - numeric coding embedded in language.

"to research, hack and extort at least 17 companies."

No surprise - ‘AI’ has a Gematria value of 17 (Reverse Reduction)

and...

Most intriguing, however, is the 37 / 73 numeric coding of ‘Claude Code’

According to the blog post, one of Anthropic’s periodic reports on threats, the operation began with the hacker convincing Claude Code.......to identify companies vulnerable to attack.

- which points to Genesis 1:1 of the Hebrew Bible; explained here (13 minutes):

https://www.youtube.com/watch?v=DwCnPQchHZs

Included in the video (with a lot more), some mathematical characteristics of 37 / 73:

37 x 73 = 2701

2701 is the 73rd triangular number, and the 37th hexagonal number.

37..73 - mirror reflections of each other; 37 is the 12th prime number; 73 is the 21st prime number, 12..21 - also mirror reflections

37 x 73 = 2701; when 2701 is added to its mirrored counterpart - 1072, the result is 3773!

[mass55th]

You’re absolutely right.