Day 2: William Buchanan (Edinburgh Napier University)

Transcript

[Music] our final speaker of the block is Professor bill Buchanan professor at the Center for distributed computing networks and security at Edinburgh Napier University and the cyber Academy I think I'll sleep Raleigh has the longest title if anyone we've had on stages so far ladies and gentlemen Bill Buchanan that thank you so I am I'm the last speaker before lunch so you'll be desperate for me to finish so I am part of a what's called the block pass

identity lab and that's been built in Edinburgh so we have funding for research as the first block chain identity lab in the world which is focused on research we thinks of an identity as a core part of building a new world and a new society and fortunately few of our governments have implemented something that's scalable and is trusted by the citizen so I'm going to give you an overview of how to build a more trusted and citizen-centric future and it's up to all of us to really understand this in the context of the new world that were actually creating and I saw this core and I

forgot where I took it from but could have troubles free photography is fixing all the problems that we've set up with our internet it's a sticking-plaster SSL became sort of our protocols we added in a bit of clip tour and we call the SSL and TLS it's a sticking-plaster it basically is there to protect the tunnel what happens when either end really is unprotected so cryptography has been focused on improving our system security but we think that cryptography should be contributing to improving our

society as a pretty fundamental thing to say there that the future of our society the pair are the lowest level on cryptography just like we built the buildings of this city in the past we need to be building the future on cryptography and make sure they were building a digital trust infrastructure but that we also have human trust so data privacy resilience and sharing will be a core part of building this new world for the first times citizens will be in control so we can't really start a presentation

without bringing up GDP our it happened and it's now with us and we kind of thought well it's over what do we do now what we need to realize is that all of the IT systems virtually every single thing we've created is legacy and needs to be stopped it needs to be migrated away because it doesn't have core principles built into it at the core of GDP are as the citizens rights to own their own data have many companies here could stand up and say the citizens really can own and control and get access to their own data very few especially at larger companies

they've harvested data they've kept data and so the arm another part we need to be smarter to be able to respond to incidents 72 hours and we need to be able to detect quicker and then we need to make sure that we have the concept of sudan anima t if something can describe sudo anonymity here in good detail and I'll be extremely impressed I have no idea what that really means to me you're either anonymized or you're not sudo anonymity is a fraud model that we need to move away from quickly as a turning point and move towards anonymity if it's

acquired and then the core of everything that we should be doing in our future is encryption encryption most be our starting point for everything that we do same transactions protected proof verification wins articulate to society what it is encryption in cryptography is actually doing so that someone can look at a green padlock on a webpage and understand what that really means so that's at the core and sued anonymity really we design systems where we might keep some personal details away from our records so a patient record here has a patient identifier that's a

pseudonym and the days of paper that sat on a doctor's desk and I had a little numbered on the top and you couldn't spot that in a digital world it's not that difficult to map a pseudo identifier to an individual and once you've done it you've done it for all the records so let them move much more towards the distributed ledger approach and our data infrastructures and it isn't just to put in a pseudo an item an authorization into a ledger because that can be unpicked we need to understand how we create the salt and how we create they identify the seed of the identity

and where that is I can't believe companies still store passwords they still store hashed passwords and these days that you don't have to store a hashed version of a passwords we company still store them leaving themselves open more and more companies need to understand they need to create virtually no presence of the identity on their system and give they identified the seed back to the individual and put it on to the phone and that way there's no physically pushing the identity back to the citizen so I received

this just just the other day I thought good now I have an identity for an election register in my home city Edinburgh that's great sent as an email and do we trust email no not one single email that you receive can ever be trusted in fact you're at work probably blocks the one that is digitally silent they receive that and I clicked on the link and I went there and ups it's not secure so how can we build up trust if we send our users from an email message with a link to an

untrusted sites and then on the page that says please be assured that this email is genuine you can't make up really you can't make up they have no investment we have investment accompanies government's don't have investment and see these developers and so on you can trust this email it says on the website how crazy is that and that's about as genuine as that so this is their I receive lots of these and I keep them all I do following

them up by the way you can have a look we do follow up these things and this was sent from the d'être sacred secretary general to me to say that I've actually won some money that's as credible as that email that I received about my important thing to do with an election and just occasion wondering he does exist that's on there and I've no idea if it was from him or not perhaps I should follow up and I could get some money thank you UN I really appreciate you sending me money but what are the roadblocks the roadblocks ah that we need to build a

world with futile knowledge proof and with homomorphic encryption if you're interested you can go to the websites that were set up and find all about the cryptography involved but those are the two core things Bob sends Peggy into cave Peggy wants to prove that she knows the secret pass phrase to get through the interconnection tunnel so he lets her go in she picks her own route here B and she stands there and Bob shouts appear through B if Peggy appears through B then he's proven a

little bit that she actually knows the password she could have used B but if she's only to not be able to get to B so bob has proven that Peggy doesn't know the secret password so he keeps sending her n and more and more when he sends her and she's able to come out through the right entrance that's the core of zero knowledge proof but the future of computing is the world's homomorphic encryption and homomorphic encryption allows us to process with a process data in an encrypted way and still end up with the results that we require so Craig Gentry defines versus of having special gloves that you can operate on

the jewels in a jewel box and the workers can can work in a jewel box without actually opening up the jewel box so more and more we need to be working around these two core areas z calf you see said Newsies Zed key stat snacks to be able to provide this but look at the performance set look at the performance that it takes to actually create the proof for the zero knowledge proof massive massive overhead and this is the problem with both homomorphic encryption and and homomorphic encryption little knowledge

is there as a massive overhead and the at reproduction at the time so basically it works I can take encrypted values from Bob and Alice as a data processor can then add subtract divide multiply find things and I as a data processor cannot tell what the values were and what the result actually is that's a perfectly secure world because processing and memory is the biggest problem you can stop virtual machine you can look at its memory so memory will have all its variables it'll have keys and things like that so we have to move much more to watch this type of world and then we can do things

like finding out who's older between Bob and Alice without actually having to ask the name so the concept of Trent will disappear the content of trust what's the core of trust in Internet PKI does anybody understand digital certificates yes over here maybe few people in the industry understand PCI and digital certificates it's a fraud system the hackers now can buy let's encrypt certificates for zero dollars soand don't trust the green padlock anymore it's a flawed system because

we're still relying on these trustees current identifiers identifiers that we didn't add to their legs and what happens if they get hacked what happens if Google's private key gets leaked to the internet how low could be trust in that world so we should be able to start to be thinking about the kind of things the kind of things that we might want to ask without revealing any core information and that is a GDP our proper world so way it works is you take your values you have a secret key you can do your function get it sold back and you can decrypt it

the core of it is actually to create what called cypher bits this is like ones and zeros and then we operate on the cypher beds in a logical way just like logical circuits and then we decipher them and we convert them back to the number us again so it's almost like we need to rebuild all our computing libraries even just the add and subtract and multiply we need new libraries which will operate on all cypher boats look at any of the cipher bits and you will not be able to determine what the actual original value was she'll interested there's the clipped or behind Oh fantastic mass it's quite simple and

there that we're dealing with individual ones and zeros there's what an adder looks like if you remember maybe from school timer to the adder looks like that we build a crypto based on an L tronic circuit so the circuit becomes a mathematical generator for this and we can see companies like IBM building massive computers to be able to cope with the requirement for this type of world because processing will be key to make sure that we preserve the success of the data and then what we can do is that we can mine across our blocks and

we can put encrypted data on to the blockchain and then use homomorphic encryption to be able to do simple things like what we have to do to find out that Bob has got enough money to spend on his account but should we be revealing how much Bob has got on his account and should we be revealing who he's spending money with and the answer in GDP Aras no and until we get that problem solved we will not get our citizens to take their money from their banks and put it on to the blockchain because they're hanging up their washing for everyone to see

that key snark brings along two contributions hidden homomorphic is that I can feed values into a smart contract and the smart contract cannot tell what those values are it can quite happily operate on them in the right way and give me back the result but there's no way that the smart contract can tell what the input and the output values are it also allows for what's called blind evaluation I can use a smart contract but there is no way I can determine what the mechanics is of the smart contracts that's kept secrets to the user of the smart contract the

one system that could do this is in development and the hawk system allows for the smart contract to be held off the blockchain and still be trusted so it binds the smart contract to be running cryptid from a cryptography point of view off the blockchain but it will it doesn't take up the processing requirement required for the main blockchain an example of some cord is this for at Farrar for a smart contract so Hawk could provide us with a way that we can start to relieve the Lord on the ethereum blockchain by running smart

contracts that can run anywhere and cliff achates has been spoken about and clip the kitties should that we could have ended up with a bottleneck so the time we kind of reached the limit of what was possible and atheneum but haven't managed to push through but to fundamentally push the barrier we need to scale up even even more and the transaction speeds really repo can't be included as a proper blockchain type solution but it achieves rate

that are moving towards visa but if we need to move Athy the Armen Bitcoin up that scale to overtake visa we need new methods to be able to implement this because we've now come to bottleneck and it needs to be removed so the the solutions that are currently been proposed are around sharding and Casper so sharding is a bit like our old economies where each of the islands are able to trade and keep their own ledger and what can dependently and all we really need to do is if the island's trade with each other is that we have some transaction system that we

keep some receipt system that shows that one island has transacted with another one that's a sharding method Casper brings a method where we move away from the waste of energy of proof of work towards proof of stake and sharding works by having a global route and then you create shards whether this will work it's yet to be seen but we're now going to trust shards to look after their own ecosystem and that would work because and most of the applications would develop we create a little ecosystem where we just trade amongst that system and only occasionally do we have to jump

out and communicate with another system so the model that's proposed is this one and we're going to have a shard which will have an ID as we create them and then each shard will keep its own transactions and receipts for those transactions and then we will have an interconnect record which will actually define can go to it when the two shards communicate with each other so our receipt will be sent from one shard to another that will be processed and then

a confirmation will be sent back and then that way the charts can ease the burden on the overall blockchain and it's not really been thought through completely yet but we'll have to see if it works so as there is a development platform for laws to develop applications Kasper also tries to move away from the proof of work method which is really bonding up too much energy and moved towards a proof of stick and there and it's reckoned that the amount of work that will be required to secure the blog or throughout from 1500 ISA to only

32 so the problem just now is that we just make it too difficult too expensive for somebody to take over the network make it too expensive and it's going to cost them a lot and in an electricity but with a proof of stake we drop it down and to set it to ethers so those are the things that are now starting to to change and scale up and what we want from our house and many countries want this is that we've built up that health system which is clinician focused we revolve around clinicians and we tell us what they want and we booked appointments with them and so on

most countries want to move towards a citizen as an active commissioner and for citizens to take control of their own health and well-being in blockchain the areas of possible innovation are multiple so I can see elsewhere like health records are an obvious one but things like asset tracking be prescribing systems are almost a no-brainer in terms of that application it just seems a natural thing to put an app an automated prescription system onto a block team

and the problem with health care is that you will see someone goes through multiple paths and they will have multiple illnesses and only with a blockchain can you actually create a health records on an audit trail which can actually determine where a patient is at any given time and it is like health insurance we can start to audit and make sure the smart contracts will enact and make payments automatically so when certain milestones are are delivered in health care we can make sure that the payments actually happen but it's also important and things like the recall of

equipment so in this case half a million peacemakers were found to have a software problems and needed to be updated health many health companies many health agencies don't keep records of which peacemaker went and which patient so we could actually have an asset tracking system in health care which made sure that if we needed to we'd be able to trace where assets were and and help the patient in Pharma in clinical trials there is so much many problems with fraud that a blockchain solution would make sure we could trace through any pharma and clinical trials

so what we're working on at the present time is an is an asset tracking system for health where we can track a device from its creation to its disposal at the end and in that way you can actually understand the dynamics of how devices are used within a healthcare environment and you can quickly recall on the case where there is a problem with the software and you know exactly where all of your your devices are so just now we're not putting the core data onto the blockchain we're doing what most people will do is to put the record of the device into trusted agency and put

assigned record a thumbprint onto the blockchain not perfect because the trusted agency could not give you back the data or might delete it and so on so there are better models that would allow you to be able to set up a trust network a better solution is that devices should be created with a private key and well should be able to sign themselves onto the blockchain in this way it provides the public key and the public key can then be proven on on the blockchain unfortunately few devices are ever

created with any sort of identity and the problem is is that RFID devices don't have the computing power to be able to do complex public key encryption methods so really it's open for us but if we can create a system that starts the track assets we can understand the full cost of a device even when it leaves the hospital where does it go how does it get decommissioned and and IRA making sure that things are disposed off correctly so our vision of the future will look something like this much more citizen focused there is a possibility of personal health records

this identity is owned by this the the citizen and not by the healthcare provider the transactions which need to be public are there and then the rights to create write update and delete are there for both the the citizen and the clinician the deletion of data is our biggest problem that we need to address as an industry gdpr asks for it if somebody wants to delete that data off the blockchain what are you going to do and the only way you can do it to stay aiive trampled the encryption keys I have arrays decryption key you can have

a look you will not find it do you think lawyers understand that concept if I get access to encrypted object is that a hack or not just because I've deleted the encryption key it's going to take me a zillion years with all the resources in the world to break that is that still a data breach so more and more we need to be enacting around smart contracts and we need to make sure our ecosystem works and this will help small businesses to come in and provide small services if I were to provide a service for one hour to change bandages in my local area I can come into my ecosystem I can get paid through health tokens and so on and

this creates a much healthier economy for health care but more and more it breaks this it increases trust and increases trust because we do know the mistakes happen and they happen for all the most unfortunate of reasons and we need to address that by having verifiable or the trails that can actually enact smart contracts and understand where the patient has been and what actually happened so the deletion of data is a important part of this a smart contract so our worlds this is what we're building just now looks

like this in fact one free bus travel in Edinburgh when I'm age 60 I should walk on a bus and automatically I shouldn't have to pay a smart contract somewhere as determined has created an acid station from me and I stored it somewhere and as enacted a smart contract so when I walk in a bus I will get free transport so there needs to be a sovereign identity there needs to be a secretions that are signed by trusted entities so we need to create ecosystems which we can sign and especially in health care for certain things just as we've done in the past with pieces of

paper we need specialists to sign digitally sign artists stations just like we do with certificates and then store these and use these to act on our blocked him and our healthcare our social care and government systems should be ready and it should be well defined what those contracts are that scares a lot of legal professionals the amount of time and effort wasted with paperwork and signatures and the whole process it's extremely difficult so our world will change and will probably move towards something more like this so

here's a smart contract for social care it sounds very simple but this is just providing care for somebody to wash their dishes clean the windows and so on so upfront the smart contracts to find there's an hourly rate there is qualification levels there's a time taken we can actually measure them at a time and upfront if we can prove the the inputs and smart contract we can make the whole system work we're a center fising people here to improve the qualifications because they know the smart contract if they can provide the evidence they will get a higher rate for this things become more open and more

sustainable so we would like to build a more trustworthy system I do cryptography so I can build digital trust what's missing is the human trust element Estonia gives one of the best examples that I've seen for human trust the Estonian system for health care means that any public sector person who views one of your records the citizen is allowed to question why that has happened and that public sector individual must respond back with an assent time limit to say why they looked

at your records that is human trust in action I can provide the best little knowledge the best homomorphic encryption the best encryption doesn't matter they just hope citizens hope that's all going to happen anyway and obviously it doesn't what we need to do is to make sure that our citizens really understand that what we're building is something that's on their site so go back to this again is that photography has been focused on the sticking-plaster of our systems and it needs to build a new world a new other people to do it thank you

[Applause]