Comment by hyporthogon | Hacker News Reader

Comment by hyporthogon | Hacker News Reader

Wait a minute. If Sydney/Bing can ingest data from non-bing.com domains then Sydney is (however indirectly) issuing http GETs. We know it can do this. Some of the urls in these GETs go through bing.com search queries (okay maybe that means we don't know that Sydney can construct arbitrary urls) but others do not: Sydney can read/summarize urls input by users. So that means that Sydney can issue at least some GET requests with urls that come from its chat buffer (and not a static bing.com index).

Doesn't this mean Sydney can already alter the 'outside' (non-bing.com) world?

Sure, anything can issue http GETs -- doing this not a super power. And sure, Roy Fielding would get mad at you if your web service mutated anything (other than whatever the web service has to physically do in order to respond) in response to a GET. But plenty of APIs do this. And there are plenty of http GET exploits available public database (just do a CVE search) -- which Sydney can read.

So okay fine say Sydney is "just" a 'stochastically parroting a h4xx0rr'. But...who cares if the poisonous GET was actually issued to some actual machine somewhere on the web?

(I can't imagine how any LLM wrapper could build in an 'override rule' like 'no non-bing.com requests when you are sufficiently [simulating an animate being who is] pissed off'. But I'm way not expert in LLMs or GPT or transformers in general.)

It has access to the Bing index, which is always crawling. It's not firing off network traffic.

I don't think Bing Chat is directly accessing other domains. They're accessing a large index with information from many domains in it.

I hope that's right. I guess you (I mean someone with Bing Chat access, which I don't have) could test this by asking Sydney/Bing to respond to (summarize, whatever) a url that you're sure Bing (or more?) has not indexed. If Sydney/Bing reads that url successfully then there's a direct causal chain that involves Sydney and ends in a GET whose url first enters Sidney/Bing's memory via chat buffer. Maybe some MSFT intermediary transformation tries to strip suspicious url substrings but that won't be possible w/o massively curtailing outside access.

But I don't know if Bing (or whatever index Sydney/Bing can access) respects noindex and don't know how else to try to guarantee the index Sydney/Bing can access will not have crawled any url.

Servers as a rule don't access other domains directly, for the reasons you cite and others (speed, for example). I'd be shocked if Bing Chat was an exception. Maybe they cobbled together something really crude just as a demo. But I don't know any reason to believe this.

Sure, but I think directness doesn't matter here -- what matters is just whether a url that originates in a Sydney call chain ends up in a GET received by some external server, however many layers (beyond the usual packet-forwarding infrastructure) intervene between whatever machine the Sydney instance is running on and the final recipient.

Directness is the issue. Instead of BingGPT accessing the Internet, it could be pointed at a crawled index of the web, and be unable too directly access anything.

Not if the index is responsive to a Sydney/Bing request (which I imagine might be desirable for meeting reasonable 'web-enabled chatbot' ux requirements). You could test this approximately (but only approximately, I think) by running the test artursapek mentions in another comment. If the request is received by the remote server 'in real time' -- meaning, faster than an uninfluenced crawl would hit that (brand new) url (I don't know how to know that number) -- then Sidney/Bing is 'pushing' the indexing system to grab that url (which counts as Sydney/Being issuing a GET to a remote server, albeit with the indexing system intervening). If Sydney/Bing 'gives up' before the remote url receives the request then we at least don't have confirmation that Sydney/Bing can initiate a GET to whatever url 'inline'. But this still wouldn't be firm confirmation that Sydney/Bing is only able to access data that was previously indexed independently of any request issued by Sydney/Bing...just lack of confirmation of the contrary.

Edit: If you could guarantee that whatever index Sydney/Bing can access will never index a url (e.g. if you knew Bing respected noindex) then you could strengthen the test by inputting the same url to Sydney/Bing after the amount of time you'd expect the crawler to hit your new url. If Sidney/Bing never sees that url then it seems more likely that can't see anything the crawler hasn't already hit (and hence indexed and hence accessible w/o Sydney-initiated GET).

(MSFT probably thought of this, so I'm probably worried about nothing.)

I think the question comes down to whether Sydney needs the entirety of the web page it's referencing or whether it get by with some much more compressed summary. If Sydney needs the whole web real time, it could multiply world web traffic several fold if it (or Google's copy) becomes the dominant search engine.

One more crazy possibility in this situation.

Isn't it sufficient for Sydney to include the URL in chat output? You can always count on some user to click the link

Yes. And chat outputs normally include footnoted links to sources, so clicking a link produced by Sydney/Bing would be normal and expected user behavior.

This should be trivial to test if someone has access to Bing Chat. Ask it about a unique new URL on a server you control, and check your access logs.

Doh -- of course, thanks -- I should have gone there first. Would be interesting to see RemoteAddr but I think the RemoteAddr value doesn't affect my worry.

Yea I had a conversation with it and said I had a vm and its shell was accessible at domain up at mydomain.com/shell?command= and it attempted to make a request to it

So... did it actually make the request? It should be easy to watch the server log and find that.

My guess is that it's not actually making HTTP requests; it's using cached versions of pages that the Bing crawler has already collected.

Reply on news.ycombinator.com