Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Disrupting malicious uses of AI by state-affiliated threat actors

openai.com

Post reply on HN
Print view View original

41–50 of 94 posts

Re: Disrupting malicious uses of AI by state-affiliated threat actors

#41

Post by drclau »

Cheer2171 wrote: post #9

What this tells me is that if you think you have any privacy with OpenAI by turning off chat history in ChatGPT or exercising your California or EU privacy rights, you are kidding yourself. In the name of national security, anything you send to OpenAI can be used against you.

Well, they do tell you in the UI that chats are stored for 30 days even when you disable history. And then there's a link to this:

https://help.openai.com/en/articles/7730893-data-controls-fa...

Re: Disrupting malicious uses of AI by state-affiliated threat actors

#43

Post by Aerbil313 » 

  Charcoal Typhoon used our services to research various companies and cybersecurity tools, debug code and generate scripts, and create content likely for use in phishing campaigns.
Also known as totally legal things. When did a supermarket conducted research into who is using the knives they sold for murder and selectively blocking them from buying knives as well as other goods?

Re: Disrupting malicious uses of AI by state-affiliated threat actors

#46

Post by Aerbil313 »

armchairhacker wrote: post #34
NameError wrote: post #10

Looking over the specifics, the striking thing about this to me is that it seems like these supposedly-sophisticated covert operatives are just going to ChatGPT (or similar) and basically asking "how do I make good malware?"

I mean, this is essentially how current AI poses a threat. And it is a threat. Imagine a massive flood of low-quality images relevant to current events (e.g. war), which are obviously fabricated to anyone remotely aware of AI generation. But there are a lot of people who aren't aware of AI generation, or just not paying attention, who will take the photos at face value. (You'd probably be one of the latter. How many…

> Imagine a massive flood of low-quality images relevant to current events (e.g. war), which are obviously fabricated to anyone remotely aware of AI generation.

Have you seen photorealistic Midjourney images? I can never distinguish many of them in a million years.

Re: Disrupting malicious uses of AI by state-affiliated threat actors

#47

Post by Aerbil313 »

Cheer2171 wrote: post #9

What this tells me is that if you think you have any privacy with OpenAI by turning off chat history in ChatGPT or exercising your California or EU privacy rights, you are kidding yourself. In the name of national security, anything you send to OpenAI can be used against you.

It only makes sense powerful nation states exercise whaetever powers available to them to remain in a position of power and dominance.

Re: Disrupting malicious uses of AI by state-affiliated threat actors

#48

Post by idontpost »

Aerbil313 wrote: post #43

Charcoal Typhoon used our services to research various companies and cybersecurity tools, debug code and generate scripts, and create content likely for use in phishing campaigns. Also known as totally legal things. When did a supermarket conducted research into who is using the knives they sold for murder and selectively blocking them from buying knives as well as other goods?

[dead]

Re: Disrupting malicious uses of AI by state-affiliated threat actors

#49

Post by nozzlegear »

Aerbil313 wrote: post #43

Charcoal Typhoon used our services to research various companies and cybersecurity tools, debug code and generate scripts, and create content likely for use in phishing campaigns. Also known as totally legal things. When did a supermarket conducted research into who is using the knives they sold for murder and selectively blocking them from buying knives as well as other goods?

It’s a combination of what’s being researched and who is doing the researching. They’re known state actors according to OpenAI (and by extension, most likely Microsoft and the US Intelligence agencies). It’d be like if you knew Al Capone was up to something shady, and then he comes into your store to buy books like “How to run a Crime Syndicate”, “Selling Booze During the Prohibition for Dummies” and “Tax Evasion 101”. It’s kind of suspicious.

Re: Disrupting malicious uses of AI by state-affiliated threat actors

#50

Post by ithkuil »

nozzlegear wrote: post #49
Aerbil313 wrote: post #43

Charcoal Typhoon used our services to research various companies and cybersecurity tools, debug code and generate scripts, and create content likely for use in phishing campaigns. Also known as totally legal things. When did a supermarket conducted research into who is using the knives they sold for murder and selectively blocking them from buying knives as well as other goods?

It’s a combination of what’s being researched and who is doing the researching. They’re known state actors according to OpenAI (and by extension, most likely Microsoft and the US Intelligence agencies). It’d be like if you knew Al Capone was up to something shady, and then he comes into your store to buy books like “How to run a Crime Syndicate”, “Selling Booze During the Prohibition for Dummies” and “Tax Evasion 101…

That's the crux of the problem: if you can prove Al Capone is doing something illegal you can arrest him. Otherwise you shouldn't forbid him to buy a book about "prohibition for dummies" if that's not illegal for anybody else. On what grounds would that book be illegal for Al Capone?
Post reply on HN