Machine Learning is Needed to Detect Money Laundering

[rmilly]

Blockchain forensic companies such as Ciphertrace and Elliptic are developing deep learning techniques to monitor money laundering on the Bitcoin ledger. Now researchers at MIT and IBM are working to apply deep learning techniques to detect the ledger even further.

As Dave Jevans CEO of Ciphertrace has mentioned "Tracking funds through the Blockchain system requires the most advanced computer science and cybercrime knowhow."

Curious to see how these companies will plan to work with governments and if they will bridge the gap between regulators and the crypto company.

https://www.theblockcrypto.com/2019/08/02/blockchain-analytics-startup-elliptic-mit-researchers-collaborate-to-detect-money-laundering-in-bitcoin-using-deep-learning-techniques/

[1715077283]

1715077283

[1715077283]

1715077283

[bitsurfer2014]

I don't tolerate money laundering schemes inside crypto networks such as Bitcoin and I welcome this initiative to use deep learning to detect such illicit activities yet I would like to point out that fraudstera will always find a way to circumvent it thus I hope it  could be really effective and foolproof once implemented. Imho.

[meanwords]

I don't tolerate money laundering schemes inside crypto networks such as Bitcoin and I welcome this initiative to use deep learning to detect such illicit activities yet I would like to point out that fraudstera will always find a way to circumvent it thus I hope it  could be really effective and foolproof once implemented. Imho.

We'll see how well this machine learning is but it's going to have a real hard time telling from change addresses, tumbling that's not for illicit means, exchanges and what not from real laundering. I mean how the hell could one ever tell if they take Bitcoin transfer it to monero (they don't know the monero address) and private send it to an exchange, get dash or some other coin trade it for monero again private send to an exchange and buy new bitcoins.

True, it would take them a lot of trials and errors just to teach the AI to distinguish between the addresses, it would take years. Not to mention tracking them and be able to see if it's money laundering or not. I'm curious if they will be able to pull this off properly.

[Deallove9]

It has always good to work toward making the community great and stopping the launders , the regulators are good to main a script community and technology by protecting users and the tech,

[avikz]

A statement from the article,

Bitcoin’s association with illicit activities like drug trafficking and hacks is often brought up by critics, but detecting such activities in Bitcoin is much easier compared to physical cash because of its transparent and pseudo-anonymous nature.

Money laundering and using of crcypto in illegal trades, have been two major pain points for many countries. Government servants often mention such terms while addressing the media when the regulation of cryptos come into the picture. But it is actually much easier to track such payments in a blockchain environment, thanks to the "distributed ledger technology"! I just hope that Ciphertrace and Elliptic will be able to bring some kind of vigilance around it so that we can win back the trust from the government and convince them to regulate and adopt cryptos in the main ecosystem!

Majority of the Asian countries are still skeptical about cryptos! Unless they are assured that cryptos are much easier to track than conventional money, they will continue to maintain such negative stance about cryptos! Just think about the endless possibilities of cryptocurrencies if these are adopted and circulated within the mainstream economy! 

[pooya87]

these are companies that are created to sell certain services related to blockchain analysis so they keep coming up with weird things like this claiming to be innovating new ways of analyzing the blockchain while all they do is repeating what has been done for many years and also were known ways of linking transactions together for a long time.

[Kakmakr]

I wonder how surveillance companies will be able to track transactions that were done through Mixer services.  I know I read somewhere that it becomes easier to track bitcoins, when large amounts of coins are pushed through Mixer services, but I do not know if this is possible with smaller transactions.

I think they will use machine learning to spot transaction trends that might indicate that money laundering are done, because bigger money laundering operations are shifting a lot of money on a constant bases. 

When the can break Mixer services, all financial privacy goes up in flames. 

[timerland]

This sounds like a really weird article. It's like they think they can do anything with AI and machines these days, which is just not the case.

Now, blockchain analytics startup Elliptic has partnered with researchers from Massachusetts Institute of Technology (MIT) and tech giant IBM to apply deep learning techniques to analyze over 200,000 bitcoin transactions as part of an effort to detect illicit activities such as money laundering and ransomware.

This is still pretty interesting though. 200,000 transactions, what will be the difference between them? Will some be flagged as illegal darknet transactions, and some be flagged as regular ones for the machine to be able to learn from it? This article just seems like putting 2 random buzzwords together to make a big deal.

When the can break Mixer services, all financial privacy goes up in flames.  

That might be crazy. I've seen machines be really good at learning patterns and track bank and other transactions via it, this seems like a very possible option, and a very gloomy future for us if it's true.

[eternalgloom]

It has always good to work toward making the community great and stopping the launders , the regulators are good to main a script community and technology by protecting users and the tech,

And how are users protected by this exactly?
This is just another attempt to shove even more regulation into the cryptocurrency sphere.

I'm sure there will be ways to circumvent this and the regular users will be the ones really affected by this.
Just so they can send more automated letters, reminding people to pay taxes on their crypto holdings.

[Nadziratel]

Money laundering is something that has been going on for many years. This has nothing to do with the crypto. On the contrary, I think that if a controlled decentralization is presented in the right ways, a more transparent economic life can be sustained by cryptocurrencies. It's ridiculous to be treated as if there were no economic crimes like money laundering and theft in the past (before cryptocurrencies)!

[seoincorporation]

Their real fight is versus the mixing services, is fun to see how they use artificial intelligence to try to decode those mixing services. But i think they will not be able to do it. But at least they are allowed to try it, since blockchain is public information anyone can play with the data as they want. But i don't like much the idea of private business tracking all transactions.

[hatshepsut93]

I wonder how that will work. Machine learning requires a ton of good data as an input, so the network can learn the patterns and then see them in any data, but where did these companies get a huge list of money laundering transactions? And how can we be sure that they don't screw things up and start labeling as money launderers innocent people who use mixers to get more privacy?

[alexcopper]

I wonder how that will work. Machine learning requires a ton of good data as an input, so the network can learn the patterns and then see them in any data, but where did these companies get a huge list of money laundering transactions? And how can we be sure that they don't screw things up and start labeling as money launderers innocent people who use mixers to get more privacy?

That is a good point. Would it make sense that these machine learning capabilities only be available to public blockchains say like Bitcoin or Ethereum versus private ones? I would be interested in understanding how they separate innocent people versus those that are using mixers to get more privacy. My guess is movements in and out of exchanges and movements of money across countries quickly.

[copoyes]

very good technological development
but a little worried whether there will be a problem or bad news about bitcoin if the machine is ready, but it is very good
because there must be money laundering through bitcoin

[13abyknight]

This comes in as good news as this move will give authorities a better shot at tracking down output addresses from tumbling services. The only thing I'm waiting to see is how they will deal with time delay between input and output spends even when the amount of BTC would be high and big transactions keep popping up rapidly into the pool.
Either ways, this is good stuff as this will reduce money laundering in the form of Bitcoin over time and launderers will have second thoughts of using mixers.

[barbara44]

This comes in as good news as this move will give authorities a better shot at tracking down output addresses from tumbling services. The only thing I'm waiting to see is how they will deal with time delay between input and output spends even when the amount of BTC would be high and big transactions keep popping up rapidly into the pool.
Either ways, this is good stuff as this will reduce money laundering in the form of Bitcoin over time and launderers will have second thoughts of using mixers.

What I would have prefer that they deal more on right now is how to make sure that the system of cryptocurrency is free and void of scammers because this is the one that is really tarnishing the image of cryptocurrency and not money laundering, for now, I am not even sure that people use it that much to launder money, maybe because those that are into money laundering does not know about the cryptocurrency yet.

The one presently at hand now which is quite dangerous to the survival of crypto is that of scamming, especially through ICO, this has to stop because it is getting too much, just last week here, about 6 of my friends were all scammed and now they are so much discouraged about cryptocurrency, we don’t need people leaving crypto now.

[Artemis3]

I wonder how that will work. Machine learning requires a ton of good data as an input, so the network can learn the patterns and then see them in any data, but where did these companies get a huge list of money laundering transactions? And how can we be sure that they don't screw things up and start labeling as money launderers innocent people who use mixers to get more privacy?

Well this could only be a tool to assist detecting suspicious activity at best, but it isn't going to be accurate, especially at first. For example a a flag would be the fact that it used a Mixing service. I'm not against them, and people using them aren't necessarily money laundering, but it increases suspicious level and its one of the factors the AI will take into consideration.

But, i don't think it will help much. It might help catch a few lazy money launderers, but the good ones will be able to do it undetected anyway. Chances for false positives is too high in my opinion.

[yoseph]

I don't tolerate money laundering schemes inside crypto networks such as Bitcoin and I welcome this initiative to use deep learning to detect such illicit activities yet I would like to point out that fraudstera will always find a way to circumvent it thus I hope it  could be really effective and foolproof once implemented. Imho.

True, it's high time something like this come into play to help put a stop to moey laundering activities which tends to give cryptos such negatives report by the media though there are a lot of positives which they don't report.

[btctaipei]

well...

You have Value Laundering (transactions involving btc to tangable goods / services) mixed with Money to bitcoin conversions (and vice versa).

What selective indicators can you attribute to a money laundering output? You need to build a classifier with sufficient good training sets to effectively apply kernel method and relavent Markov chain to the raw blockchain transaction data.   This is where I would begin and put my attention to

[rmilly]

I don't know if anyone has recently saw the AML report recently published by Ciphertrace but I found it interesting in seeing the current landscape of money laundering and exactly why we should look into ways to help track. I have a feeling that mixers are going to become illegal especially after the Bestmixer incidence. We're going to see much more governments start to step in

https://ciphertrace.com/q2-2019-cryptocurrency-anti-money-laundering-report/