Hacker News Reader

Comment by neilv | Hacker News Reader

> Threat actors buy vulnerabilities that fit into existing business processes

Isn't there a market for this? For example, "Reveal who is behind this account that's criticizing our sketchy company/government, so we can neutralize them".

I'll also argue there's separate incentives, than the market value to threat actors... Although a violent stalker of an online personality might not be a lucrative market for a zero-day exploit for this "threat actor" market, the vulnerability is still a liability (and ethical) risk for the company that could negligently disclose the identity of target to violent stalker.

IMHO, if you're paying well a gazillion Leetcode performance artists, to churn out massive amounts of code with imperfect attention to security, then you should also pay well the people who help you catch and fix their gazillion mistakes, before bad things happens.

View on HN

You are imagining a market that doesn’t exist.

First there are only very few gobs/companies that are sketchy enough to do this - and for those a huge number of non-anonymous people exist with huge reach that are very critical for years. If such a market would exist they would assassinate all those first - you don’t need the email if you have the face, voice, and name - since that is not happening they just don’t care that much about it.

There’s 100% an active market for this, and I think tptacek is simply wrong on this point (the others are valid)

The likes of Cambridge Analytica didn’t go away, they exist and absolutely go hunting for data like this.

The ability to map between different identifiers and pieces of content on the internet is central to so many things - why do you think adtech tries to join so many datapoints? Let alone things like influence campaigns for political purposes.

I’m not talking about assasination plots, but more mundane data mining. This is why so much effort in the EU has gone into preventing companies from joining data sources across products - that’s embedded in DMA

There's an easy way to put your money where your mouth is here. Just offer $11k for this or similar vulnerabilities out of your own pocket, and then resell them. If there really is a large and active market for this at higher dollar values, you'll make a killing!

Sure is funny there's nobody doing that despite so many people being so dead certain there's an active market.

If I did, would you know?

And if I did, it wouldn’t stop people from doing co-ordinated disclosure either, would it? Same with high end exploits - some folks do co-ord disclosure because it feels good and is great for your CV; others sell gray market and we generally have no idea what’s being traded.

(With the exception of say, zerodium or 0xcharlie’s various talks)

Which of "0xcharlie's various talks" addresses the likelihood of your being able sell a web authz information leak bug on a Google site for bitcoin?

Sure, but do adtech companies buy vulnerabilities in web services to advance their mission? Wouldn't that risk running foul of e.g. the Computer Fraud and Abuse Act?

You don‘t need to sell the vulnerability to them, or even tell them the vulnerability is there. Just set up an API and bill them by the query.

This ignores tptacek's points in the top-level post.

> [...] a bug that Google can kill instantaneously, that has effectively no half-life once discovered, and whose exploitation will generate reliable telemetry from the target.

You can't set up unmask-as-a-service because it's going to take you longer to get clients than it will take Google to shut down your exploit.

Yes, but:

1. It can still take a while before Google finds out

2. You can log every mapping you got in the meanwhile, then keep selling the ones you already have

Edit: although probably most of your business will be over when word gets out that your data isn’t exactly legal (which your clients have understood from the start, of course; they could just plead ignorance)

People keep talking about this as if there's a 0% chance of being caught if you do this?.

So let's suppose that you did set up the service like this. Can you even make 10 K? What are your odds of getting caught? How much do you value not being in prison and/or having to hire a lawyer to get you out of there?

I'd take the 10k every time.

I’d take the 10k, too, but I think it’s possible to pull this off without getting caught.

It’s a lot more work, of course, but you can scrape some top youtubers first as it seems relatively easy. If you can pull this off you can then try and figure out how to legitimize your offering – I won’t go into details here, for obvious reasons, but now that you have something valuable on your hands it makes sense to spend some time/money on selling that.

You’re talking about this as if there aren’t other countries who actively infiltrate power infrastructure and for whom this is the most low risk mild attack (if you can call it that)

I’m not speaking theoretically, which I suspect most on this thread are.

Okay, which state actor is going to buy this for $100,000? How are you going to sell it to them? What's the risk of getting caught?

Even if someone on telegram was telling me that Russia would buy this information for $100,000, I think I would reach out to Google and "settle" for $10k.

I’ve seen a light version of this, where a “marketing data” company was scraping baby shower gift registry pages and selling the data to an infant formula company in the US.

The scraping was def in violation of the EULAs. Product data is one thing, but I believe this group was combining it with other sources and selling the identities and context as a bundle.

An API is too much work. Grab the addresses for the top 100,000 YouTubers and sell that csv on the dark web.

What happens when the first to buy the CSV starts selling it themselves?

That’s not a new problem with selling info on dark web marketplaces. if you're interested in learning more, here are a couple of books you might enjoy:

"The Dark Net” – Jamie Bartlett “We Are Anonymous” – Parmy Olson “Future Crimes” – Marc Goodman “Kingpin” – Kevin Poulsen

I think you've missed my point. I know data brokers exist. Does there exist today a data broker that functions in whole or in significant part buy acquiring vulnerabilities and exploiting them to collect data? He's a more concise way to frame my argument: if you're imagining yourself to be the first person to sell a particular kind of vulnerability to, then your customer is imaginary.

Yeah, I think this is valid. “I’m confident I can find someone who will buy this” vs “I’ll message grugq”, roughly?

My feeling is that if he were still paying attention on HN he'd probably back me up on this stuff (if not, I'd be thrilled for him to come set me straight).

i think what's being conflated here is that there are reasonably buyers for this kind of vulnerability but there's no market in the truest sense. I think a correctly connected individual could well sell this vuln to a state actor or a contractor to one; but the ecosystem of bug sales to these parties has no aggregate appetite for them, thus, there is nothing driving the price up. People in the market for cyberweapons want point and shoot vulns that have broad usage beyond a specific server for a specific company or parts for them, and ones that will last beyond a single corporation patching something. They are willing to pay such big $$$ for this that the whole market is optimized for it. The power players here would much rather buy a gun and shoot the lock off a door than a specialised set of picks that work for that lock in that building.

The only real market (that I can see) are shady data aggregators. Governments just file subpoenas, and abusive megacorps can file lawsuits (all the anti-SLAPP statues in the world can't prevent your Google account from being unmasked and having to pay for a lawyer). There is a limited market in the form of internet addicts who want to harass people for kicks (since finding an email gives them another route to do that with), but it's a small one. These people also tend to be entitled pricks, so they're not a very good customer base to have.

> then you should also pay well the people who help you catch and fix their gazillion mistakes before bad things happens.

You missed their point about the business model of the security researchers here: their business model is finding a large number of small value vulnerabilities. Those who are good at this are very very good at this.

My company has a bug bounty program and some of the researchers participating in it make double or more my salary off of our program, but we never pay out more than this for a single report. And it's not like we're particularly vulnerable, we just get a steady stream of very small issues and we pay accordingly.

They're right: I was talking about the business models at the buyers that these vulnerabilities have to slot into. The point I'm making is: there already has to be an operating business that's doing this for a vulnerability to be salable at all. If there isn't one, you're not selling a vulnerability, you're helping plan a heist.

Right, I'm only responding to the last part where they imply to these researchers are not well paid. I'm saying that on an hourly basis or monthly basis $10k a vulnerability is actually quite a good payout when you have a surface area as large as Google's to explore and know what you're doing.

Their last paragraph shows that they didn't understand your paragraph here:

> For people who make their nut finding these kinds of bugs, the business strategy is to get good at finding lots of them. It's not like iOS exploit development, where you might sink months into a single reliable exploit.

> Their last paragraph shows that they didn't understand

I think I understood. The last paragraph of mine that you cite was speaking of the creator of the bugs, not the discoverer.

The liable party should be investing reasonably towards non-negligence. (Especially in the context of spending billions of dollars each year on oft-misaligned headcount that's creating many of these liabilities.)

I'm not talking about the company optimizing for the minimal amount they think they can get away with paying to try to cover their butt. Nor am I talking about how white/gray-hat researchers adapt viable small businesses to that reality.

Yeah, _should_ but businesses make money and not reporting and using the vulnerability in any other way is illegal, so they get to set the price as they're the only buyer. They know this.

Reply on news.ycombinator.com