Cyber Security, Privacy and Trust PhD at the University of Edinburgh

yeah so um this session is about the cyber security privacy and Trust PhD program so this is one of uh many PHD programs offered by the school of informatics in the University of Edinburgh and our program is run by the um cyber security privacy and Trust group within the School of informatics so we have one of the uh largest such groups in the UK research groups working in cyber security topics um and I'm going to give you an introduction to the program what it offers uh talk a little bit about the application process um and then give a chance for questions and then in the second half talk a little bit about research areas so as we're quite a large group there are many different research areas and I can't hope to cover them all um and we might also run out of time depending on how many questions in the first part but I'll try to at least introduce some of the uh research interesting research past research that we've done and uh some future ideas as well okay so um cassia have I got control yes yeah yeah yeah bottom bottom of my screen yeah so um as I said we're we're quite a large group in terms of uh the way these things have measured it means we have about 20 um staff who are uh uh uh teaching staff and academics um as well as about 30 postdocs and 25 phds in progress that's that's out of a much larger body in the school of informatics overall so um we also have some interesting connections with other schools other departments in the University so particularly people working on topics like um uh social sciences and law um so uh the university as a whole um but particularly thanks to our research group is recognized as a UK government's recognized Center of Excellence for research uh an acsr there are about 20 of those universities across the UK that have uh reached a certain bar which says that their research uh is large enough and good enough that um we can be considered a center of excellence so it's a good place to come um and also actually our research outputs are measured uh our research papers um in a uh a ranking which works for computer science called computer science rankings by measuring outputs in some of the top conferences um and we're definitely within the top five across the whole of Europe for that and by some measures in the in the uh actually number one um depending on which group of topics you you look at um so we have a sort of particular uh approach in our group which is to Foster these multidisciplinary collaborations and I'll mention some of them uh a little bit bit later on but um but also as well as as well as going broad we also go deep so um we look uh a lot of our research looks at um fundamental topics so theoretical aspects so um yeah so you can think of it as being about foundations with applications so we have some good connections with um industry as well as some industry sponsored research the activity that takes place uh right right here or in the building next door so we're uh sort of embedding some of our research in uh current practice as well um and then by being placed in the capital of Scotland we have a sort of nice ecosystem around us um based in the capital city and some of the activities that are going on in the region uh but also within Scotland and then within the UK and then uh internationally as well uh so um uh I've just got a few photos here of people I can't introduce you to all of them but but uh encourage you to look at our web pages which has connections to each of the the people that have mentioned here and their research interests and when I get to the research topics a little bit later there'll be some photos you might have trouble matching the photos to the names if you if you have trouble doing that for a particular research area then you can always uh email me afterwards but um but good to have a look at our uh list of PhD topic ideas uh on our web pages anyway um and I just want to give a little bit more of a ground on um what the program is about so the fundamentals here um perhaps you already know these already but uh I just want to give a few thoughts and um an aspect on uh security proving trust so really these are problems that cut across nowadays nearly all aspects of how we interact with computers um and we think about um privacy to begin with um sorry I'm just checking my slide so privacy is a little bit of um uh sorry this this should have come as the second slide let me let me talk about cyber security first then I'll go back to privacy so um cyber security you you you might have seen before um uh this traditional CIA Triad so CIA stands for confidentiality integrity and availability these are sort of three most common properties of computer security that more more or less everybody Lear about when they first take a first computer security course so confidentiality is about the um uh about preventing the unauthorized learning of information so not just information that um that should be denied but you shouldn't be able to actually learn the information by some other means if you're not uh authorized to do so um Integrity is about the unauthorized protecting against the unauthorized alteration of data or information um and availability is about maintaining in the uh access to data so you can see the the sort of uh little Graphics there giving you that that kind of indication of how uh how those different accesses are uh denied or should be allowed um and then if we think about privacy um privacy is a little bit from our point of view um as computer security researchers we have a particular take on privacy so privacy is uh a different property it's about partial sharing of information between parties um so it's often often people think about privacy as necessarily being concerned with personal information but it might also be about information uh concerning organizations or or groups of individuals and the um the essence here is that where some some information may be shared but other information might not be so for example I might be happy to reveal to a survey uh the first half of my postcode address to say roughly which city and which area I live in but I might not want to reveal all of my postcode because that would narrow down who I am to just a handful of houses for example so um so we're seeing the rise of privacy enhancing Technologies nowadays which help to manage this partial sharing of information in very carefully uh managed ways and then if we think about uh trust so these are the three parts of our program security privacy and Trust so trust is really uh something that underlies underlies some of these other properties because when we're talking about um sharing of information and whether sharing of information should be authorized or not really we've got in the background some idea of uh being able to decide on a security policy for sharing and also being able to decide on who um participants um might be so having some form of uh authorization uh authentication and then authentication linked to author Iz ation so you see there are um Notions of trust and identity which are fundamental um so we call that cyber trust um and then there are two two sort of uh strands to trust that we're particularly investigating at Edinburgh one is the use of blockchains and distributed Ledger technology which provide new solutions for forming uh trust as a consensus mechanism so um by by some distributed agreement protocols uh and the use of blockchain um and then another aspect of trust which we're very interested in is trust in systems themselves so trust in software and hardware and how we can assure that by uh using techniques like verification of software or verification of Hardware devices uh or uh and also conveying the fact that some devices may be trusted um uh in in a distributed system so that's known as attestation so hopefully that gives you a a sort of background for what most of the PHD topics are in cyber security privacy and trust and then here's a list of some of our specific research themes so um I'll read these through very quickly there's a bit more in the second half of the talk on some of these themes but I won't have time to cover them all so we're looking at topics in Ai and security certainly um you can read this in both ways so um we have ai systems being applied to cyber security so um so that enables new um new ways of providing security that might be driven to AI but we also have very important questions about trust in Ai and security of AI systems so we want to know whether AI systems are secure for deployment then I've mentioned already um the distributed Notions of security and uh data science that might be applied and particularly in the application of blockchains and distributed ledgers so uh I mentioned that before we have a group who work on that but also work on more General ideas in cryptography so public um uh public key cryptography and uh also a little bit of symmetric key cryptography but not uh not so much research on that here um but a range of topics within public key cryptography um related techniques um and uh we have some work on uh networking and ideas for future future networks including networks of devices as well as um uh as well as the infrastructure for networking itself um we have work on Hardware devices so looking at things like Internet of Things devices and how do we ensure privacy and security for those devices um we have more General work looking at security and reliability in systems and how we might develop new systems using um forms of uh secure design and also um also take into account new new emerging facilities which are being provided by newer Hardware that provide new forms of trust essentially um then we have strands work in protocol and program verification so I mentioned verification already so we're interested in verification not not just in software uh itself but also in the design of security protocols and how one establishing security Properties by a sequence of exchanges of messages and um that's a that's an area that's uh been very rich for computer security research because very short protocols can be used to establish trust between two parties that may not may or may not have communicated before but they can be um surprisingly complicated to get right and have subtle flaws in their design not only uh in their implementation um then we have a strand working on Quantum Computing and particular Notions of quantum cyber security so I'll say a little bit more about that uh later but there's a new CDT program actually in the area of quantum which would be a good place to apply if you're interested in Quantum specifically um and then finally we have a a strand with several researchers working in Notions of soot technical aspects of uh computer systems and their data so uh it's important to recognize that um you know security is not just uh lying in the technology it depends on how uh people interact with technology as well so that brings in things like human factors as well as the socioeconomic environment too so things like uh uh law and assessment of risk um so let me tell you a little bit now about our uh PhD program so um hopefully already uh if you're watching this talk you have an idea about what a PhD is but the basic idea for a PhD is it's a training that you can undertake to become an academic researcher okay it might be something that you uh you undertake because you want to understand uh topic very deeply and contribute to research in that area and that might be the main takeaway for you it might be something that's very useful for your future employment but if one of your future ideas is to become an academic researcher uh like like the group here then um this is also absolutely an essential thing that you need to have before undertaking that career path so in general we hope you develop an all around knowledge of the discipline um and also develop excuse me a second there's some noise from outside close my window uh I'm not sure if you could hear it um we hope you developing all around knowledge of the discipline I we hope you develop some uh knowledge in some specific Advanced Techniques which will be in your specialist area of study so a PhD goes uh uh very deep on a very specialist uh topic um but at the same time we we cover a broad range in all of our PHD programs of transferable skills which will be useful to you more generally so those are things like the ability to uh present your work clearly to others and uh to convey you know deep um deep research results but in more accessible ways for example um and I would say that there's one overriding requirement that you should have before you consider embarking on any PhD program and that's to um have a passion for your research topic so we expect you to be you know really driven by your interest in the area and wanting to learn more and that's the most natural thing that we have from I would say all of our PhD candidates um so what do you actually do during a PhD so basically you'll carry out independent research so the research sometimes can be collaborative but the one of the main requirements to be awarded a PhD is that you have to be able to demonstrate that you've done something uh yourself you've made a major contribution yourself and that's what the uh the result of the pH will be so you produce these original contributions to knowledge in the chosen area and that those might be made manifest by writing and Publishing academic papers uh sometimes there are other forms of output including things like uh results that might get taken forward in Industry um and they might be published in other forms um but in general um in general we do expect publication and publication of results would happen in uh academic workshops and conferences usually so you will work um under guidance of uh uh PhD supervisors so and the particular pattern of supervisor is is fairly flexible but usually there will be uh at least two supervisors and there may be uh a co-s supervising Arrangement or a primary and secondary supervisor that's really uh set up during the start of the PHD and it depends on the the topic and the person really um and sometimes as I mentioned we have collaborations with industry partners there are some industry Partners who fund some of our phds um and uh there'll be the opportunity to engage with sometimes industrial supervisors who take an interest in phds um and uh sometimes uh undertake internships actually spending time with an industrial partner um so the final outcome that you're working towards is a PhD thesis so a written uh a written document that's uh you know more or less a a book so quite a substantial piece of work um and then you'll need to defend that PhD thesis in or oral examination which happens after you've submitted the the PHD and the orial exam um uh for us in the UK is um it's not too onerous it's not a public event it's a private event that takes part with uh an internal examiner and an external examiner who are subject experts uh who will um read read your PhD beforehand and then um conduct the oral examination asking you questions so that happens right at the end um but the PHD uh is always assessed according to regulations that the examiners are asked to consider and the main requirement there is that there should be a new and significant contribution to knowledge so you have to be advancing the the stateof the-art you're not only uh getting to understand what the state ofth art is but you're pushing forward you're coming out with new results um and then uh if that's all successful you'll get awarded um a PhD a doctoral degree um and then hopefully that will be useful to you in the next stage of your career path whether it's being an academic or uh maybe an industrial researcher or perhaps an engineer an entrepreneur uh a cyber security Visionary whatever you may go on to do uh lots of exciting things uh previous students have gone on to do um just to give you an idea about uh what happens between those two stages so um uh in the first year of PhD study so it's a the the program is a three-year program I should say so um so there are three to three and a half years of study before before you're expected to submit the PHD um so the first year of our program is considered probationary and this means that you'll be assessed at the end of the year primarily on uh a thesis proposal and the literature review so those are two documents that you produce uh along the way during your first year um and those documents should establish that you've um got to grips with the research area and you're able to um explain the the current stateof the art um and moreover you've got a good plan for the remaining time of your PhD to advance that state-ofthe-art so in the first year you you will be starting uh research anyway but um maybe most of the research will be uh in the second and third year so also during the first year you might have some particular uh training needs it might be that you decide that you would need to uh get a little bit more background in a particular sub area of your PhD it might be you need to understand a little bit more of a particular mathematical subject or you want to liaz with people in another school to understand more about um social or legal aspects um and those training needs as as well as any um any sort of uh transferable skills training that you might want to develop as well so those will always be offered to you throughout the program but um but it's a good idea during your first year to work with your supervisor to identify opportunities and needs there so um so uh yeah and those thing might be things like um how to do research writing or how to present well those kind of things um and also you won't be doing this just on your own and with your supervisor but you'll be uh sort of immersed in the research environment of our school which is really incredibly Rich so because we're such a large department overall we're the largest in the UK with um um uh I I think getting on for 500 PhD students in total you'll um you'll have plenty of opportunities to go to research seminars not only seminars in security privacy and trust but seminars in other topics that might be relevant so there might be Topics in uh robotics or in AI which have security uh connections that you're interested to attend and really there are so many interesting seminars every day I there aren't enough hours for me in the day to go to all the things that I'm interested in but um but it's really uh an amazing opportunity here to to engage in that um then there'll be some specific things which um you want to join in on so certain kinds of uh graduate lectures that might be useful for you to get that additional background um and then also depending on on your particular PhD topic there would be the opportunity to go to maybe specific conferences or uh things like summer schools which are offered uh to students internationally often n nationally as part of the acsr network there'll be opportunities like this but also internationally there are uh good Summer Schools where you get the chance to meet top researchers from around the world and we usually support students uh who'd like to attend those those kind of events yeah and um the subsequent years after the first year then so progress through the degree um after you've uh successfully presented your thesis proposal and had that uh approved there'll be progress report at the end of every subsequent year so that's managed by the school of informatics and then there'll be a similar um it's it's deliberately set up to be a little bit like the oral examination so we would have a a panel to review both your thesis proposal at the end of the first year and then the progress reports at the end of the second and third year um and um your thesis submission will be at the end of the third year roughly but it might be after three and a half years or so um and typically the funding packages work for um three and a half years um and if you want to find out more for information of the format of a PhD at University of Edinburgh and there's a link you can follow there um and should be very ailable to you in the slides afterwards um something about how to apply so um my colleague Daniel Woods is here as the PHD selector for this program um so Daniel would you like to present the slide yeah wait let me close yeah so my role is to read your applications and assess the quality of the applications both the research proposal that CV and match you with potential supervisors so what you will need to do is submit you know normal things like a CV some of your transcripts and for some people English language certificates but the main I guess difficult creative part is the research proposal so you know to do this I think the best thing you can do is read through the list of Supervisors see what kind of research areas are interesting and reach out to those supervisors because we are not expecting you to formulate a perfect research proposal without speaking to us and actually if you speak to a potential supervisor I actually think there's more chance that you will together find a research topic um so you know th this is kind of the high level I would really encourage you to nominate a supervisor on on your form um because you can apply without nominating a supervisor by just proposing a topic and then I will try to match you with the most relevant supervisor um but it's it's much better if you can nominate that person and yeah if for the information on the specific you know documents you need to send there's a link here that will help with that and I think there's one one more slide yeah so I guess the the main question you guys will have is what funding is available and I would say there's there's two Roots here the first route is to get General funding from the school um and the idea there is it's not linked to the topic you propose just any kind of computer science topic can be considered for this funding which is General school funding to get that funding you need to apply by the 25th of November for the first round and I would encourage you to do that it's not long now 12 days but that's where most of that funding can be assigned and essentially what will happen is your proposal will be compared to not only applications for cyber security privacy and trust but also NLP systems uh robotics so all of the phc proposals and phc applicants will be compared and we will offer funding to the top students so for that approach you know you can propose any project and the most important thing is the project proposal is strong that's where I would recommend speaking to a potential supervisor to refine it and also your proposal is sorry your profile is strong and it matches that so that is you know the most free kind of free way to do it because you can work on any topic but that is challenging like there's not so much funding available for that route the other route you can go down is the UK centers for doctoral training if you're not familiar with the UK system this is somehow between a normal UK PhD which has basically no required training and the US system where you need to do two sometimes three years of training UK cdts it's more like nine months of training so it's somehow in between the exams aren't as strict as us qualifying exams um but the UK funding counil councils have directed more funding towards these cdts and the problem I guess is we don't right now have a cyber security private privacy and Trust CDT so if you wanted to go down this route you would have to look at the cdts that exist so there's one in natural language processing there's one in Quantum Technologies there's one in machine learning systems and you would potentially look at a cyber security privacy trust topic that is linked to those so something like the cyber security implications of quantum Technologies would be a good candidate Maybe security of natural language processing systems large language models would be another good system or how we can apply machine learning um to IM improve cyber security you know th those kinds of topics could potentially be funded by one of these cdts you would have to do this and it it does vary a little bit between the cdts but you would have to do some training some courses but I actually I did a CDT when I started my PhD and I actually found the courses really useful because they are you know they're often very quick they're they're taught Often by potential supervisors so they're much more relevant to the research you will do um yeah and there is more funding available for these cdts so if you have interests in Quantum natural language processing machine learning systems and there might be a few more I think there's one related to space um so space cyber security is a topic that you might be interested in then I would encourage you to look at the cdts and there will be a different application cycle you won't apply to uh the cyber security privacy and Trust PhD instead you will apply to those cdts but you might work with a cyber security privacy trust faculty member to kind of prepare your application and with the goal of working with them um so there's quite a few examples here and um one to say is I'm not an expert in all these areas so I'll uh explain them as best I can or just try and give you an impression about them but if I spark some interest in uh a particular area then have a look at our list of PhD topics and you probably can guess the one most closely connected with the research that I've talked about and contact the supervisor concern um and by the way if you when you reach out to a a potential supervisor try to be very specific in what you write don't just Spam lots of people and um if they think they're the wrong person they would uh if it's a nice message they might refer you on to one of the colleagues who's more suited so um and I can't cover research of all of our uh colleagues here so first of all let me uh mention some work in human factors um and uh as I mentioned this is uh this topic area is you know dealing with socio technical aspects of systems so how can we support people in their use of technology and not uh not not blame people for security problems but have a uh overall secure environment um and one of our uh faculty members our staff members working in this area is uh Jing Lee so actually he's he's got his full contact details here just because he he gave this slide in a nice ready to use form um and he's working on topics in user Centric privacy and security um and dealing with things like uh AI agents so how can we be sure that AI agents we might have ai agents that are even sort of somehow um agents that are uh very close to ourselves maybe with the latest kind of um uh AI technology so they're acting uh as us um and even be appear to be us but how can we make sure that they do that in a way that preserves the security and privacy goals that we might have for ourselves you know not revealing they might know a lot of information about us uh an online avatar or agent that access us but um but we might don't want them to reveal too much um so uh so that's an important uh topic um and look at that link between users and technology and how we design both privacy and security at once um another area is in trust uh and there we're looking at work in database Technologies and in um programming languages and techniques for managing the FL flow of data through systems so understanding where data has come from is often very important to defining trust so if we want to have trust in some uh data set for example that's been used for training an AI model we might want to know where all the pieces have come from and be sure that we trust the sources of all the different pieces that go together to make up a a large data set so being certain of the data's accuracy as well as knowing where it's come from if we can um so there are some quite deep um theoretical techniques and methods for for doing that I mean the program at the top looks a bit uh basic but the idea um the idea is if you have a much more sophisticated function that might compute with a a bunch of inputs how can you uh track the data that might be computed so if you sum two numbers together you know you have the provence comes from the the combined Provence comes from the two sources individually so it's related to problems um that have been studied for a long time in computer security of information flow but goes a little bit further um and then there are interesting questions also on the sort of flip side so if we have systems that change the nature of computing to track where information has come from in this this format then that can be useful so that we can find out Providence but it also can introduce new privacy problems because it might be that we we track too much so how can we uh measure that um that kind of thing um another aspect of trust uh this is is an older project but quite interesting I think is one that was conducted together with um uh design informatics here so a group that works on um design in the School of Art in in Edinburgh and here uh the idea was to actually try to monitor um the circular economy and understand how value is uh is maintained by reuse and Recycling and reselling uh Goods so um so that's more to do with um aspect again aspects of trust and aspects of information flow and digital traceability which um can have security as well as privacy implications um another another project with the School of Design which is uh nice to talk about was a uh um a coffee machine that was uh equipped with a Bitcoin interface so that you could purchase your coffee with Bitcoin um this was very early in the days of um uh seeing any kind of Bitcoin interfaces for physical devices but um the sort of the interesting aspects here was not so much that we we could put a device onto a coffee machine and pay by Bitcoin but actually that it enabled tracing some new and enabling some new forms of choice so we could uh give people a choice of the beans that they might choose uh and then a particular coffee machine might become uh more adapted to ordering beans from one place over another according to user's choice and those choices could be related to ethical decisions of about um the uh the the beans that were being purchased or or or flavor decisions those kind of things um different topic area verification and modeling how do we know for certain that the Technologies we create are secure what does secure even mean so there I mentioned in the um the first part that we've uh had a lot of work in this area in the foundations for program and protocol verification so designing formal Logics that you use to reason about programming languages but also reason about protocols um and then verifying the proofs so often cryptographers work on handwritten proofs so they're like uh dense mathematical arguments uh and those are fine but sometimes they can have mistakes in them so converting those uh handwritten proofs into proof on machine um makes the proof a little bit more robust and testable and also allows us to test changes so if you have a very long mathematical proof it's quite difficult to uh assess the impact of every single uh assumption in the proof but once we have these mechanized proofs um that's uh that becomes enabl so uh enabled so some of our group are looking at tools and techniques for doing that um and uh yeah uh one topic area that I've worked in is the use of uh uh modeling systems so describing systems abstractly um and then verifying aspects of the design and we've done that for Notions of um Notions of security for mobile apps and App Stores um and we've also been doing that for um looking at interactions between models for Safety and Security Systems and we're looking at safety I'm looking with a new PhD student starting this year at interactions between security and privacy on uh sort of risk-based graphical models so if you've seen a attack tree the picture there on the bottom right will be uh familiar um as well as verifying AI systems which I'll say something more about here's a little picture just to show you um the give you a sort of hint of what was going on in some of that earlier research so one of the mobile applications that's uh types of mobile application that's been um uh really had a bad reputation in the world the Android world has been um a whole series of flashlight applications which are designed to just turn on the lights on the phone so that you could see and back in the old days um lots of these applications were uh distributed to many users um and they did turn on the lights but they also uh had a lot of other Behavior which was sometimes stealing people's personal information and um by doing some program analysis of the applications we were able to look at the application there and you see on the right there's enormous amount of code to do with advertising but actually the functional behavior of the flashlight is very uh is very small in there so um so it's quite interesting to be able to take apart um uh applications or uh uh or even malicious code and understand it in that way and try and um see where where insecure actions may take place um yeah Quantum cyber secur as I mentioned is a subtopic within our Quantum group and I won't say much more about that because there has been a separate briefing for the quantum um Quantum informatics cdts so have a look at that briefing or ask cashia to send you uh send you uh a link if you weren't um weren't at that briefing to uh to find out more but there's a lot of interesting work there so basically in the uh the applications of quantum Computing more specifically to security problems and even the way that even having some little Quantum elements somewhere on the network uh can enable better security Properties by making use of that element so it doesn't mean that we all need to have quantum computers on our desk there might be just some Quantum interaction some Quantum Cloud elements so that's quite exciting uh and a lot of um uh research funding being um being put into that in the UK um we've also got a long strand a long lift strand of work on EV voting um which is um coming out of the cryptography group there and uh a bunch of um uh bunch of techniques and protocols for conducting voting electronically that sometimes you know they provide at least as good guarantees as you would hope from physical vote voting with um pencil and paper but actually some more properties can be guaranteed as well which is uh which is one of the appealing things so so sometimes there's horror stories about uh EV voting may be going wrong but actually it has the potential to do better than current voting which is quite exciting so you're able to actually check whether your vote has been counted for example through the clever techniques uh We've also looked at verifying the security of hardare we Bitcoin wallets um that's been a nice piece of work by MTO there and um yeah and in general um privacy enhancing Technologies so we've got several researchers who are working on privacy enhancing Technologies how do we enable people organizations even governments to protect uh privacy um and well there's been lots of examples where I I mean this is very difficult problem this is a just a slide to sort of suggest the problem there's an Infamous case where uh or organizations or a CLE several cases on this slide where organizations have re released data for the purpose of data science research and then only realized later that the data is subject to uh linkability attacks so one data set can be connected to another data set and then reveal information that the uh people re releasing the data didn't uh expect so one of the first big examples of that was Netflix that released anonymized viewing data and then researchers connected that to data sets that were publicly available from crawling IMDb and they were able to De anonymize some of their data which is pretty scary you know so that's a fundamental problem um and then there are anonymity networks like the tour Network which are designed to preserve anonymity um in browsing or uh uh navigating the web um and we've got researchers who look into the properties there to make sure that tour does what it should do and doesn't leak more information than it should um so quite a lot of going work even though tour is you know deployed and in use um another aspect is uh statistical models for security um and privacy and uh how much information is necessarily revealed when we do release some statistical information and how can we add noise in a way that uh doesn't reveal too much so new privacy definitions um we also have uh a strand of work and I expect this is going to increase over the coming years in responsible Ai and AI for privacy and we're probably going to see more overlap between security computer security researchers and researchers who are working on responsible and safe AI but Nadine is one of uh our uh Team there who's working on pets and other techniques to um to make sure that AI systems are deployed in responsible ways and also built and trained in ways that uh preserve the privacy of people involved um I've mentioned already the work on blockchain and distributed Ledger that's quite a significant subgroup within our uh group and uh we have a strong connection with industry there so we have an embedded industry working on uh properties of blockchain and um looking at Future distributed Technologies for the cloud like multi-party computation and fully homomorphic encryption uh so that's quite exciting um and then in hardware and networks um uh we have work on AI for security and security of AI and um I've been working there on some techniques for uh looking anomaly detection for um deciding if Intruders might be breaking into a network um and similar on the other side is uh some work on verification of AI systems using using a tool that we've developed in a project in the last few years actually that's quite exciting um there's also work on things like fitb devices or uh well this is older work now but uh but in general still interested in these kind of problems with um Hardware security of uh devices and um uh devices at the end and the edges of networks and how those can be integrated with mechanisms for uh attestation or just doing the best that we can if we don't have full Trust on remote uh devices uh that's quite an interesting one um we have a large project on uh rigorous engineering for secure and uh trustworthy systems that I don't have time to go into but it's supported by a lot of um work in the uh UK and then finally and this is some slides from Daniel actually Daniel but I'll just talk to them quickly uh we have this important question of Law and public policy so defining the rules and how they should be interpreted so there there are interesting questions because of J different jurisdictions how applications have to behave differently in different places um so uh and Daniel's area more generally is answering questions about managing cyber security and uh risk and how those are modeled and dealt with um I'm afraid we're out of time so I have to wrap up now and um if you have any lingering questions then um there are uh Pages yeah so NE for next steps you might need to like to look at these Pages here um for virtual visits of the city or chatting to students um and then for followup questions as cassia mentioned at the beginning there's a general email future students ed. add. or you're welcome to write to me directly if you'd like so I guess future students has the advantage that it will it maybe anonymize your questions if they're passed on to us but but for that you pay for the delay in passing on the questions so um but some of the questions might be better handled by the central admissions rather than um uh course team so that's uh that's everything from me thank you thank you very much for your attention

2024-12-28 16:18

Other news