Sei pays out $2M bug bounty
usmannkhan.com
usmannkhan.com
Good companies use bounties as yet another security layer - after doing everything else, add a bug bounty!
Almost all crypto bug bounties run through Immunefi. [1] There are lots of > one million dollar bounties. You can see SEI's current bounty page here.[2] The company I work (a different company) for has a one million dollar bounty listed on immunefi.com and median response time of six hours.
Jacking a database and trying to sell it on a DLS or dark web is a massive process.
Contrast that with cryptocurrency where a bunch of VC money pumped up a market for you to launder the proceeds and the protocols are intentionally designed not to have antifraud protections. Ransomware was possible a decade earlier but the profitability went up massively once it became easy to launder millions rather than hundreds of dollars.
This explains reliable, stealthy, zero-interaction full-chain iOS vulnerabilities, which fit into every intelligence, military, and law enforcement business process pin-compatibly. It explains browser vulnerabilities and ATO vectors.
And it also approximates the market for blockchain vulnerabilities: if the exploit is "literally transfer untraceable cash from victims to buyer", lots and lots of criminal organizations already have that business process; you probably simplify their existing repeatable process.
Blockchain vulnerabilities thus have a very credible market. As bonus: the work of discovering and POC'ing these vulnerabilities may be gnarly, but the engineering required to exploit them at scale probably isn't. It doesn't take months of R&D to make the exploit "reliable", it generates straight cash until it dies (and probably has a half-life measured in minutes), and so on.
Every lucrative class of vulnerability has some kind of story like this; they all fit into some existing, very clearly stated demand.
We get into trouble trying to generalize. All the markets are very specific; they're all sui generis. Most vulnerabilities are worth zero. There are mobile OS RCEs that are probably worth zero!
Aren't there quite a few cases already where attackers stealing funds from smart contracts were considered just that: thieves. And where their "code is law" defense didn't amuse the judge?
IIRC we recently even saw two sent to jail for manipulating smart contract prices: it's not even clear they used a bug in a smart contract.
I already posted it but Uncle Sam cannot have it both ways: if Uncle Sam asks people making money with cryptocurrencies to pay taxes, Uncle Sam goes after those who steal from the taxpayers. And... Oh boy, does Uncle Sam tax gains.
They certainly don’t accept “locks are law” or “finders keepers.”
The MEV and Sandwicher attackers are legal, increase the transaction costs for everyone, skim profits from everyone and annoy everyone, the exploiter of a MEV bot gets charged and convicted.
I don't have any problem with that, I've analyzed the sentiment of discussion though.
I don't think anyone got charged and said code is law. Its more about who gets charged at all.
See also all the people pissed at zelle.
For example, the national bank of Bangladesh was compromised in 2016, believed to be a well-resourced attack by North Korea, and the attacker was able to attempt to transfer $1B. That’s about as severe as it gets, but the U.S. Federal Reserve blocked 85% of the transferred funds and of the remaining funds, all of the money sent to Sri Lanka was recovered, and they were able to recover some of the funds laundered through a corrupt bank in the Philippines whose manager was subsequently charged. About $64M was laundered through casinos which were not at the time required to follow KYC.
https://www.bbc.com/news/stories-57520169
So, not great, but the losses are under 10% of the amount the hackers had access to and there’s still a chance of recovering the rest - that’s survivable with insurance and it’s basically the traditional finance world at its worst in terms of corruption & poor preparation. Compare it to cryptocurrency, where losses on that scale happen multiple times a year rather than once a decade, and the attackers have a much easier time laundering funds through the infrastructure setup for exactly that purpose. North Korea is getting over a billion dollars a year from cryptocurrency, which is much better than the tens of millions at greater risk they got here.
All I was saying was that banks have a bug bounty on their head, the next person responded that bank transactions are reversible, which isn't entirely true in all cases.
I wasn't trying to compare sizes or anything like that.
The only way you are recovering the bulk of losses if you don't notice the theft very quickly is if the amount is high enough that a prosecutor is interested and it hasn't all been withdrawn as cash yet.
I have the opposite conclusion there, crypto organization sponsored bug bounties are far more accurately valued than Web 2.0’s arbitrary adversarial bug bounties, and have attracted tons of developer talent to crypto bug bounties and the crypto ecosystem as a whole
when Apple is going to fight tooth and nail to not pay you $10,000 while the black hat government contractor will pay $1,000,000 for the same exploit, the market is saying what the real price is and its at parity with what Web 3 is paying
And yet: "Both issues were caught after the code had been audited, merged, and slated for release"
I wonder who did those audits?
I caution readers to not make rash judgements on their skill like this though. These bugs are really hard to find, and it was a minor miracle that I noticed these ones at all. I actually had a whole list of critical bugs in this codebase ready to report before the V2 upgrade was merged to master (which would put it in scope for a bounty). However the auditors managed to find every single bug on my list. I only noticed the ones that eventually made it here later, by a stroke of luck, and after I had already spent a ton of time looking at this codebase without noticing them.
did you try other things like try to get employed by the team, or consider submitting an altruistic pull request? or was the bug bounty the adequate incentive from the getgo
[1] https://bittrap.com/resources/defis-growing-pains:-as-tvl-ra...
2. From your other comments elsewhere in this thread, it sounds like you are a full-time bounty hunter, correct?
2. Well, I'm currently not employed full time and I do spend a lot of time bounty hunting. But I mix it in with other things as well, like competitive security reviews on https://sherlock.xyz or https://cantina.xyz and private contracted security reviews.
How you find those? Or this type of work finds you based on your activity on competitive security review sites?
One way to do this is to show some chops on the competition sites and then move to one of the organized freelance firms like Spearbit or yAudit. In doing all of these things you'll inevitably meet more people, build a specialty, get some reputation, etc.
There's an unofficial project that tracks bounty programs, you can see the change here: https://github.com/infosec-us-team/Immunefi-Bug-Bounty-Progr...
2. How extensive is your background in networking, blockchain programming and pen testing?
3. How many other bounties did you commit recon time to before the two successful disclosures?
2. I am a very experienced security researcher/pentester/whatever we want to call it, specifically in the blockchain niche. I'm OK at the other stuff (reversing, cryptography, web, mobile, etc). Networking probably alright? I'm comfortable saying I have a good mind for security and a wide knowledge of the basics in many fields, then a very deep knowledge of a select few areas.
3. Idk, a lot! Upwards of 20 for sure.
You can check Immunefi's Bounty-Board for reference, currently paying up to $15M per find.
Another good source is rekt.news, creating post-mortems about all the DEFI-hacks and an own leaderboard, $624M for #1.
Just like a bug in a smart lock does not allow you to enter a house because "you were allowed in".
So it is $2 million x probability payment vs $100 million x probability escape without getting caught.
Even with the threat of non-payment, not sure I could ever feel at ease with a multimillion bounty hanging over my head.
These chains are created by startups with VC money, they are not going to hire hitmans.
The more likely concern is that someone will sell you out to any of the numerous governments who feel you wronged them. Leading to decades of life in prison.
It's not so much the projects themselves who are a threat, but the thousands (?) of random individuals whose value is stolen.
I am not making a judgement about this specific case.
Here's a simplified hypothetical example to help you understand the legal nuance: I offer all of my money to the first person that can solve 5x5, and I errantly believe that it's a difficult problem to solve.
This bounty prize is the equivalent of finding a Chrome zero day bug or an iPhone zero day RCE jailbreak. There are lots of >$1M bug bounties in crypto.
The question is, would you rather target Chrome/Safari or iPhones and find and chain-up 5 - 10 zero days for $1M+ or target crypto projects instead for $2M per project?
You're really missing out.
Having the iPhone bug and the accompanying conference talk and blog post will allow you get hired by nearly any good security or tech company. No one cares about blockchain bugs except other crypto companies. When I and a bunch of other coinbase engineers were looking for jobs we were looked down at for even working in crypto. And weren’t even in the blockchain team! Just regular engineers.
I myself have dedicated a couple of months to testing gnosis and curve that each have $2 million bounties but turned up short. Last year I switched to a ML based fuzzing research and was able to speak at defcon and got crazy offers after publication.
I would find what is most like your problem domain and dig in :).
Vendor bounties for these kinds of vulnerabilities are going to tend to be sharply lower than this crypto bounty, which was for a directly monetizable vulnerability. But there's a lot going into that vendor bounty price point.
This is actually why "proof of stake" blockchains are fundamentally flawed. They only make sense if the value of the system is denominated in the currency of the system. It's self referential and prone to negative feedback loops. They are secure because the token is expensive, the token is expensive because it provides a secure platform. Short the token, take a loan out, compromise the security, tank the value, profit. All the mechanisms to prevent that are built into the system, like delaying the validator pool entry, but the only real backstop is a hard fork and spinning up a new copy.
It seems like it might be worth the gamble of taking 3-6 months off work to discover a bug of that size.
https://blog.sei.io/bug-bounty/
> Where does one go about discovering bug bounties of this size?
- SECURITY.txt for individual projects.
- https://immunefi.com for blockchain in general.
- BugCrowd and HackerOne for wider tech.
I'm an infrastructure engineer though and may not be the best person to answer.
> It seems like it might be worth the gamble of taking 3-6 months off work to discover a bug of that size.
https://www.hackerone.com/ethical-hacker/meet-six-hackers-ma...
Note: I work at a foundation for another blockchain. This doesn't affect anything I wrote above, just disclosing potential CoI.
I have always wondered why the payouts are capped at the trillion dollar corps at such low figures. It appears like $75k max and MS and $100k max at Apple. Meanwhile shady 3rd party groups will pay you 10x that, won't they?
I suppose the argument for OS makers to raise their rates might be that they are paying 10x below market rates, and the rates were set by the actual freaking market that exists.
If I was a congressional aide, I would definitely write something up about this when my boss was going to drag a Microsoft exec across the coals in public. I would imagine that billions in gov contracts are at risk for MS right now due to lax security. A $2M bug bounty could have prevented that.
Also, FAANG level salaries are pretty high for anyone involved with that type of code, right?
Intentional bug creation should probably result in firing, unless it was done under duress.
The major parties to this market are aware of each other and are calibrating against each other; Apple and Google aren't blowing this off. It's complicated and counterintuitive in a bunch of ways.
You're right though that it's a lot of risk. It's not something that most of the leaderboard works full time on, though some of us do. The immunefi homepage has a list of all the bounties on offer.
"Payouts are handled by the Sei Foundation team directly and are denominated in USD. However, payments are done in SEI." [1]
The other part of my comment is correct according to the various Immunefi listings. Again, I could be incorrect if they do something differently behind closed doors.
But OP was paid in USD anyway.
"
Cosmos uses go panics for error handling. Transaction runs
out of gas? panic. Try to spend more coins than you have?
panic. Invalid inputs? panic.
...
For safety, later on the panic was removed entirely.
"
You can issue a command to transfer currency from your account to somebody else's, as that is a primary use case of a cryptocurrency. There was a code path where you could send someone negative amounts of the currency and it would happily pay them a negative amount of currency and charge you a negative amount of currency, thus transferring their account balance to your against their will.
There were several transfer paths and I think not all of them were vulnerable, but only one has to be. There's a bit of indirection that made it somewhat less obvious than my description makes it sound, though it amounts to the same thing in the end.
This is a bug I remember from the Apple II game "Taipan" (in which you play an 1800s opium-and-silk trader in East Asia). You could borrow negative amounts of money from a lender who charges extremely high interest. As a result, the lender would quickly end up owing you tremendous sums, without your having to do anything else. Wikipedia mentions this:
> Note: A bug in the original game allows the player to overpay the moneylender, acquiring "negative debt". This "negative debt" will accumulate interest very quickly, and will count towards the player's net worth. As the game's vocabulary of number words ends at "trillion", this can cause the game to display garbage instead of the player's correct net worth. This has been fixed in the online "for browsers" version of the game.
well i guess anything can be a currency but its too misleading even though that was by design.
if its designed to be a stock then should be called so. poker chips? in game currency? money laundering token? reward points? purchase receipt? jpeg? just think it would help
A stock would be a 'tokenized equity', in the same way there's 'tokenised real estate', 'tokenised metals', 'tokenised bonds', whatever the real world asset is.
'In-game currency' is indeed used by gaming people, since that was their term from before blockchain.