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

GitHub is investigating unauthorized access to their internal repositories

twitter.com

Post reply on HN
Print view View original

351–359 of 359 posts

Re: GitHub is investigating unauthorized access to their internal repositories

#351

Post by hard_times »

port11 wrote: post #350
hard_times wrote: post #349

Earlier quoted context omitted.

Whatever findings they come up with, must be formalized to avoid non-determinism. Draft an integration test of every finding. The malicious extension makes calls to haxx0r.net? Draft a case in your integration test that intercepts this.

But why will LLMs get this right when Web-of-Trust and blacklists didn’t? For a long time we’ve had different heuristics to detect abuse, and it’s always been a losing battle. E.g. an extension that sends requests to an IP. Do you block all network access? IP ranges? Well, we’ve had firewalls for ages, hackers still craft successful vectors.

I'd say it's different this time, because LLM can form and substantiate opinions of their own, on their own. You could actually assume you have the full source code of any binary for it to analyze, since in theory everything is decompilable, and you could then also diff between releases to save 99% tokens.

Re: GitHub is investigating unauthorized access to their internal repositories

#352

Post by nottorp »

mrgoldenbrown wrote: post #331
nottorp wrote: post #262

Earlier quoted context omitted.

Can that even be done? Even if there are knobs you can turn to disable auto updates, does that cover everything that decides to change your software behind your back?

On vscode you can control autoupdates. (Still trying to find a way to control zed's overeager updating)

For vscode or for any extension as well?

Are there misbehaving extensions with their own mechanisms?

Re: GitHub is investigating unauthorized access to their internal repositories

#353

Post by insanitybit »

saagarjha wrote: post #325
insanitybit wrote: post #290

Earlier quoted context omitted.

1. We've seen LLMs detect existing supply chain attacks when pointed at malicious install scripts. This is direct, empirical support for my position. 2. We have a long history of using heuristic technologies to detect attacks. We can infer that other heuristic technologies can be combined in a successful manner. 3. Shortcomings of LLMs are directly addressed by removing attacker controlled information from the input,…

I don’t deny that LLMs can detect some attacks. I just don’t think they can be made to do so reliably.

I think it's reliable enough and cheap enough that it's worthwhile.

Re: GitHub is investigating unauthorized access to their internal repositories

#355

Post by chris_money202 »

hard_times wrote: post #351
port11 wrote: post #350

Earlier quoted context omitted.

But why will LLMs get this right when Web-of-Trust and blacklists didn’t? For a long time we’ve had different heuristics to detect abuse, and it’s always been a losing battle. E.g. an extension that sends requests to an IP. Do you block all network access? IP ranges? Well, we’ve had firewalls for ages, hackers still craft successful vectors.

I'd say it's different this time, because LLM can form and substantiate opinions of their own, on their own. You could actually assume you have the full source code of any binary for it to analyze, since in theory everything is decompilable, and you could then also diff between releases to save 99% tokens.

Think this is more an argument against them in security than for them. You don't want the LLM forming the opinion that something is secure when it is not because it has "microsoft" in the name of a binary or something else. There are a lot of ways to mis-direct their reasoning and it's almost impossible to control that 100%. A lot of AI providers are learning from the mistakes, so implies something is going to get hacked once in some unique way and then LLMs won't make the mistake again with 99% certainty.

Re: GitHub is investigating unauthorized access to their internal repositories

#356

Post by hard_times »

chris_money202 wrote: post #355
hard_times wrote: post #351

Earlier quoted context omitted.

I'd say it's different this time, because LLM can form and substantiate opinions of their own, on their own. You could actually assume you have the full source code of any binary for it to analyze, since in theory everything is decompilable, and you could then also diff between releases to save 99% tokens.

Think this is more an argument against them in security than for them. You don't want the LLM forming the opinion that something is secure when it is not because it has "microsoft" in the name of a binary or something else. There are a lot of ways to mis-direct their reasoning and it's almost impossible to control that 100%. A lot of AI providers are learning from the mistakes, so implies something is going to get ha…

Let's strip all strings then before analysis? to prevent prompt injections and such.

Re: GitHub is investigating unauthorized access to their internal repositories

#357

Post by djray »

nextaccountic wrote: post #341
_ache_ wrote: post #338

Earlier quoted context omitted.

What is the hell is muslims-of-github.tar.gz ?! I guess nebula is the new PRISM. What is `newly-packaged-malware` ?! I'm sorry for Sophie.

> What is the hell is muslims-of-github.tar.gz ?! Is Github keeping a list of Muslims in their platform? That's horrifying

Large organisations have advocacy, support and friendship groups for many underrepresented or protected groups, such as those for religions, mental health issues, physical disabilities etc. I imagine this is simply one of those internal group's organisational repos.

Re: GitHub is investigating unauthorized access to their internal repositories

#358

Post by jlaternman »

DANmode wrote: post #264
Y-bar wrote: post #247

Earlier quoted context omitted.

I’m in a location where Apple Maps is significantly better than Google’s. So I’m unsure if you mean ”it’s Apple Maps meme bad” or if you just mean ”it’s rather meh, could be better, could be worse”.

Apple Maps used to direct people off of bridges and into ditches and stuff. It’s a swell experience, now, but, the “meme” comes directly from reality.

And Google Maps literally did something very similar to me once, just a few years ago. Told me straight ahead when there was a sharp hairpin obscured by overhead bridge (literal mapping issue in unusual motorway adjacent road). Caused a crash with minor injuries I got back up and walked away from (on two wheels, would have been fatal if I didn't brake so well, or didn't get off the road fast enough, a large truck came round the corner). Takeaway is "never make driving decisions based on what the screen shows." There is no platform worth trusting more than your eyes on the road ahead.

Re: GitHub is investigating unauthorized access to their internal repositories

#359

Post by the_real_cher »

mustardo wrote: post #296
the_real_cher wrote: post #292

Earlier quoted context omitted.

Microsoft’s GitHub was compromised when a Microsoft developer using Microsoft VSCode installed a rogue extension from Microsoft’s VSCode extension library, which is moderated and hosted by Microsoft. via: news.ycombinator.com/item?id=48204312

Built with packages hosted on Microslop's NPM

Absolutely disgusting
Post reply on HN