96
$\begingroup$

In a recent press release issued by Treadwell Stanton DuPont, the claim is made that their research laboratories have successfully broken all 64 rounds of the SHA256 hashing algorithm. They further claim that they achieved this milestone a year ago (late 2018).

"While we have successfully broken all 64 rounds of pre-image resistance," said Seiijiro Takamoto, Treadwell Stanton DuPont's director of newly formed Hardware Engineering Division, "it is not our intention to bring down Bitcoin, break SSL/TLS security or crack any financial sector security whatsoever."

How credible is this claim in the wider cryptographic community?

$\endgroup$
10
  • 18
    $\begingroup$ I located the official claim. Snake oil at its finest! $\endgroup$
    – fgrieu
    Commented Sep 11, 2019 at 15:45
  • 4
    $\begingroup$ Here's some additional commentary from the folks at Soylent News. They're of the opinion (like myself) that it lacks any kind of credibility since TSDP did not first release a paper to the Journal of Cryptography, nor did they provide a collection of non-trivial collisions for examination. Oh, and Hacker News have also damned it. I'll marked this as answered forthwith. $\endgroup$
    – Gary
    Commented Sep 11, 2019 at 15:48
  • 14
    $\begingroup$ @Gary Rowe: don't count on that ("no-one will fall for it"). History proves that people fall for fake claims, no matter how incredible. My favorites are ADE 651 and avions renifleurs. Get-rich-quick schemes work - for the perpetrators. $\endgroup$
    – fgrieu
    Commented Sep 11, 2019 at 16:02
  • 2
    $\begingroup$ Actually, someone who can generate second preimages for arbitrary messages can hijack any existing bitcoin (by taking a signed transaction by that owner (public key), and creating a second transaction for which the same signature is valid; that is, has the same SHA-256 hash). $\endgroup$
    – poncho
    Commented Sep 11, 2019 at 16:26
  • 2
    $\begingroup$ Next thing to figure out: what is TSDP pumping and dumping, and then beat him to the punch. Also see DEFCON 17: Stealing Profits from Spammers or: How I learned to Stop Worrying and Love the Spam. $\endgroup$
    – user10496
    Commented Sep 12, 2019 at 0:20

4 Answers 4

153
$\begingroup$

[Update, 2019-09-12: Upon being exposed, and/or after taking advantage of market manipulation by their fraudulent announcements, the scam artists of Treadwell Stanton DuPont have retracted their claim, with more meaningless waffling technobabble about ‘laboratory equipment flaws’ and ‘ambiguous results’—yet they are still selling the same snake oil! I leave the commentary below as a record of their fraud, including magic market prediction oracles they continue to hawk as well.]

[Update, 2019-09-18: Now the scam artists of Treadwell Stanton DuPont are just straight up asking for donations, perhaps because the world pointed out they don't actually have anything to sell but they still want to take your money and haven't found a productive way to contribute to society in exchange for it.]


How credible is this claim in the wider cryptographic community?

It is balderdash by scam artists trying to bamboozle people into buying fintech snake oil, using meaningless technobabble (plus quantum!), a virtual office address on Wall Street, and a slick-looking web site with sleek gadget photos and impossible claims about market prediction.

There is no merit to this claim. If they actually had this capacity, they could easily demonstrate it by showing a SHA-256 collision, or a message whose SHA-256 hash is all bits zero. Since they are running an elaborate fraud, rather than doing cryptography, obviously they cannot do this. But for fun, and to make an indelible record on the web for the chumps who at least have the good sense to search the company's name, let's take a look at these scam artists.


From the press release, archived in case anyone tries to memory-hole it to evade prosecution for fraud.

"While we have successfully broken all 64 rounds of pre-image resistance," said Seiijiro Takamoto, Treadwell Stanton DuPont's director of newly formed Hardware Engineering Division, "it is not our intention to bring down Bitcoin, break SSL/TLS security or crack any financial sector security whatsoever."

Seiijiro Takamoto does not appear to exist on the web outside the fraudulent press release issued by ‘Treadwell Stanton DuPont’.

Up to now, it was thought impossible to use the output of the hash function to reconstruct its given input.

They don't seem to be clear on what a preimage attack does. (If the input is longer than the output, what a preimage attack finds is almost certainly not the original input but another one with the same output.)

In accordance with its disclosure policy, Treadwell Stanton DuPont won't be saying exactly how they did it — because once the proof-of-concept is out, anyone with enough computing power will be able to produce a SHA-256 collision, rendering the algorithm both insecure and obsolete.

They can't keep their story straight about whether it's a preimage attack or a collision attack.


From their web site, archived:

While the exploit won’t be detailed here, it does take quite some computing power to achieve, overlapping into the quantum computing realm. So much, in fact, that combinatorial optimization is used, along with a fully connected/coupled 32,768-bit scale, 256-bit gradation solution space, which translates to a maximum of 1.158x1077 gradations of large-scale, precision parallel processing.

wat

Remember: this is not brute force.

False. This is a brute force assault on hapless victims by a barrage of technobabble.

It is not our intention to bring down Bitcoin, which is why only 25 PHAEDRA A3497 and 25 PHAEDRA A3487 Compute Nodes will ever be constructed. Even though preliminary design, manufacturing specification and initial construction work has already begun, it will likely be 12 months before the first compute nodes are delivered to clients, worldwide.

The fraudsters are creating the illusion of scarcity to scare chumps into giving them money as soon as possible, and setting a deadline that gives them plenty of time to walk off with the money and find another scam.

PHAEDRA A3497 will start shipping around mid-October 2020 at a rate of one per week. PHAEDRA A3487 will follow suit starting around mid-November 2020. Shipped on a first come, first served basis (S&H not included), if you have secured a place amongst any of the two lists (10 BTC and 6 BTC respectively), you can still climb to a higher rank by submitting or resubmitting an accumulative higher bid.

It's first-come-first-served, except if you pay more you can skip the queue! But you have to pay in BTC, it seems. I'm sure this is for principled reasons about evil central payment processors like ‘Visa’ who could be capriciously subverted by the judicial overreach of a court order about fraud, and certainly not because BTC transactions are irreversible.

Timeline:

  1. Founding. Treadwell Stanton DuPont is created to research and develop stock market prediction algorithms using artificial intelligence neural networks.

Curious that they waited until 2017 to register their domain name, nine years after their scam was founded.


Their main web page, amazingly, manages to be worse:

There's only one proven method to beat the stock market: accurate, precise and timely world-class stocks and securities forecasting. It's like having an oracle at your beck and call. Like clockwork, know what's going to happen beforehand, every single time and take the guesswork —and risk— out of wise stock market investing.

In principle there could be a cryptanalytic attack on SHA-256, but the market will be deterministically predictable only when birds like me hang out with airborne pigs.

Here's another product with their name on it, the NEBULA6 SSE Stock and Securities Forecasting Neural Net platform, sold—as a 100 USD brochure—on eBay:

If you're a sophisticated investor, hear this: Even your initial investment appreciates over time! Treadwell Stanton DuPont accounts are scarce and highly sought-after, and only a few dozens are available (it's free to register). Their value will SKYROCKET in the next, upcoming months, with many standing in-line for a chance to grab an account. ABSOLUTELY NOTHING TO LOOSE! This is a NO-RISK, ZERO LOSS, FAIL-SAFE, FOOL-PROOF investment! EASY TO USE: Even a total newbie who knows next to nothing of the stock market can get up and running in LESS THAN 15 minutes! TSD does not sell securities - they offer extremely accurate stock forecast charts with a degree of precision previously unheard-of, that has until now been the domain of some of the biggest and most established financial players (aka large investment banks).

I'm ready to line up outside their virtual office at 30 Wall Street which is so exclusive anyone on the web can rent it and make my millions!

Maybe they broke into the cryptobabble fintech space because the sketchball eBay fintech predictimacator market was too smart and got wise to them too fast.

$\endgroup$
7
  • 18
    $\begingroup$ Notice how "work on searching, developing and testing practical collision and pre-image vulnerabilities on the SHA-256 hashing algorithm began back in 2016 in Treadwell Stanton DuPont's R&D facilities", when even their domain name was not yet registered. $\endgroup$
    – fgrieu
    Commented Sep 11, 2019 at 15:54
  • 6
    $\begingroup$ @fgrieu Being registered via godaddy.com is also usually not a good sign. $\endgroup$
    – towe
    Commented Sep 12, 2019 at 11:09
  • 14
    $\begingroup$ It's marvellous that their new claim is that "a hidden laboratory equipment flaw... led to incorrect, ambiguous results". What could be ambiguous about generating a hash? - either it matches, or it don't $\endgroup$ Commented Sep 12, 2019 at 21:42
  • 2
    $\begingroup$ "The fraudsters are creating the illusion of scarcity to scare chumps into giving them money as soon as possible, and setting a deadline that gives them plenty of time to walk off with the money and find another scam." It wouldn't even need that. "We've only built 25 and 25..." because "it is not our intention to bring down Bitcoin". Doesn't work that way. Sorry. Once you have a method to compromise it, then it is over. Once those machines are out people will figure out how to make more, and you won't be able to control that. IOW, were this true, everyone should be shitting their pants. $\endgroup$ Commented Sep 13, 2019 at 1:27
  • 3
    $\begingroup$ "It is balderdash by scam artists trying to bamboozle people into buying fintech snake oil" -- you made my day! $\endgroup$
    – jcaron
    Commented Sep 13, 2019 at 11:04
37
$\begingroup$

This is firmly in "put up or shut up" territory.

It's easy to prove this claim. I will provide them with a string and its hash. They must provide a collision.

They could also mine bitcoin and earn a large bounty without actually doing harm to the block chain. They could mine faster with less energy than the competition. This wouldn't hurt anyone and would provide strong evidence for their claim.

Edit: the fact that the press release is slanted toward "cryptocurrency investors and traders" makes me think this is part of some kind of scam, given that like 90% of everything in cryptocurrency is a scam.

$\endgroup$
5
  • 26
    $\begingroup$ Even easier way for them to prove their claim: publish a preimage of the all-0 hash. $\endgroup$
    – poncho
    Commented Sep 11, 2019 at 16:19
  • 4
    $\begingroup$ In all fairness, if you have a black box that produces a first preimage for SHA-256 after humming quietly for 12 hours, that would be a devastating break, but would not be useful for mining bitcoin. $\endgroup$ Commented Sep 12, 2019 at 1:09
  • 1
    $\begingroup$ @HenningMakholm: Since they only need to prove it works, use the test blockchain. It sometimes goes a day or two between mined blocks. Just set the apparent difficulty to something stupid high. $\endgroup$
    – Joshua
    Commented Sep 12, 2019 at 1:53
  • 3
    $\begingroup$ This wouldn't hurt anyone. Pardon my ignorance, but isn't Bitcoin mining a zero-sum game? $\endgroup$ Commented Sep 12, 2019 at 12:04
  • 1
    $\begingroup$ @poncho : aka the hardest bitcoin there is to mine $\endgroup$ Commented Sep 13, 2019 at 11:42
23
$\begingroup$

As of today (12th September 2019) Tradwell Stanton DuPont have retracted their claim:

enter image description here

Sorry for the image link, but the text on their site doesn't seem to be permalinkable, and I wanted a more permanent source, it says:

NEW YORK, NEW YORK, UNITED STATES, September 12, 2019

The Wall Street fintech Treadwell Stanton DuPont again broke silence today as it announced its Research & Development and Science Teams discovered a hidden laboratory equipment flaw which led to incorrect, ambiguous results during third and fourth tier trials concerning their attempt to break the SHA-256 hashing algorithm. The firm has previously announced it had succeeded silently over a year ago. The announcement aims to assure financial and security sectors the hashing algorithm is still secure.

"Rest assured all sectors can still sleep quietly tonight," said CEO Mike Wallace. "Our preliminary results in this cryptanalytic research led us to believe we were successful, but this flaw finally proved otherwise."

SHA-256, which stands for Secure Hash Algorithm – 256bit, is a one-way function that converts a text of any length into a string of 256 bits, known as a hashing function. SHA-256 is a member of the SHA-2 cryptographic hash functions designed by the NSA, with all major SSL certificate issuers using it to enable encrypted communication between a web browser and a web server, which is by all current standards deemed secure and trustworthy. With this news, it is still confirmed impossible to use the output of the hash function to reconstruct its given input, even though Treadwell Stanton DuPont's cryptanalysis did not work exactly this way. It should also be remembered that Bitcoin makes heavy use of the SHA-256 cryptographic hash function too. Crypto investors can now relax as there is no impact from the recently announced (flawed) discovery.

While the best public cryptanalysis has tried to break the hashing function since its inception in 2001, work on searching, developing and testing practical collision and pre-image vulnerabilities on the SHA-256 hashing algorithm began back in 2016 in Treadwell Stanton DuPont's R&D facilities. It was thought to have culminated 2 years later with the successful discovery of a structural weakness and the initial development of the first practical solution space of real world value by its researchers. This is not so anymore.

"We are being forced back to the drawing board," said Seiijiro Takamoto, Treadwell Stanton DuPont's director of newly formed Hardware Engineering Division. "We will be tightly testing, validating and confirming our advances with support from specially appointed cryptographers and observers before making any such announcements in the future."

In accordance with its disclosure policy, Treadwell Stanton DuPont still won't be saying exactly how their advances progressed nor reveal any of their research just yet. Cryptographers have long been predicting a collision like this would require at least 100 years before someone produced a smart enough algorithm which actually finds and exploits a structural weakness, given the mathematical laws that govern hash functions. And while it is inevitable that hash collisions should occur because the input data is potentially infinite but the output length is fixed, this prediction is still holding its ground as of today.

So, regardless of the issues raised in other answers, we can conclude SHA256 has NOT been broken by Treadwell Stanton DuPont.

$\endgroup$
12
  • 30
    $\begingroup$ Oh, so their claim is that they thought they could find preimages, but never thought to check whether what they computed were actually preimages? That sounds very believable. $\endgroup$
    – Maeher
    Commented Sep 12, 2019 at 9:24
  • 5
    $\begingroup$ Working permalink thru the marvelous archive.org $\endgroup$
    – fgrieu
    Commented Sep 12, 2019 at 12:57
  • $\begingroup$ Are you saying that we can conclude that SHA256 has not been broken by Treadwell Standton DuPont because they said so? $\endgroup$
    – Fax
    Commented Sep 13, 2019 at 13:37
  • 3
    $\begingroup$ We can conclude they did not break SHA-256 because they're obviously fraudulent scam artists. $\endgroup$ Commented Sep 14, 2019 at 1:29
  • 1
    $\begingroup$ @Fax: If I say I've been to Mars you cannot disprove it. But you certainly can conclude it is false. You introduced the word "disprove", this answer only says "conclude". $\endgroup$ Commented Sep 23, 2020 at 15:09
-9
$\begingroup$

Well, I'm not quite buying it. Just because we don't want our bitcoin investment going up in smoke doesn't mean we have to hide our heads in the dirt and pretend it's not happening. Am I the only one noticing strange details in their retraction? "We are being forced...", "...all sectors can still sleep quietly tonight", "specially appointed cryptographers and observers". Anyone notice the upper right image too? Bitcoin chart tanking just to recover with the judge gravel? Center image with the eye of providence and the dollar? Really?

$\endgroup$
4
  • 4
    $\begingroup$ Uh... what? I'm sorry, but SHA-256 is not broken. Do you know how cryptographic research works? $\endgroup$
    – forest
    Commented Sep 14, 2019 at 1:24
  • $\begingroup$ What is your basis for believing they have a viable attack against SHA-256? $\endgroup$ Commented Sep 14, 2019 at 12:28
  • 4
    $\begingroup$ This is a web site about cryptography. If you're looking for a horoscope, or numerology, or other forms of hogwash, try somewhere else. $\endgroup$ Commented Sep 15, 2019 at 15:00
  • $\begingroup$ This answer only contains suspicions. I'll quote "Extraordinary claims require extraordinary evidence". It has been noted that it would be easy to substantiate the claim. Unless anything of substance is added to this answer, it may need to be removed. $\endgroup$
    – Maarten Bodewes
    Commented Sep 19, 2019 at 8:46

Not the answer you're looking for? Browse other questions tagged or ask your own question.