What Is Post-Quantum Cryptography?
Lory Thorpe likens the task to being told the nails in a house you built over 30 years will one day disintegrate. Do you know where all the nails are?
In this episode · 15 chapters
- From 2G to 5G, and into quantum
- What healthcare should be doing now
- Harvest now, decrypt later
- Getting governments and clients to take it seriously
- Did researchers really crack RSA?
- Which cryptography the quantum threat breaks
- What NIST standardised, and why it took years
- Can you use the new algorithms today?
- Crypto agility, and knowing where your cryptography is
- Is the threat overblown?
- Why starting early is the whole game
- Cryptographic debt
- Where cryptography already sits in your day
- Is it somebody else's problem?
- What AI changes about all of this
And once it’s stolen, it’s stolen. So there’s not a lot that we can do with data that’s been stolen, which is why we need to be thinking about that data that has a longer shelf life. We need to be protecting it now.
Lory Thorpe works with the IBM Quantum team on post-quantum cryptography after decades in telecoms, from 2G to 5G. She explains why data harvested now can be decrypted later, and why public key cryptography such as RSA is the main target.
She walks through the first three algorithms NIST standardised, why crypto agility starts with knowing where your cryptography sits, and why new systems built on classical cryptography add cryptographic debt. “Everything is secure until it isn't.”
Lory Thorpe of IBM's quantum team argues on the Thinking On Paper podcast that organisations should prepare for post-quantum cryptography now, rather than ask when a quantum computer will break today's encryption. Because data stolen today can be decrypted later, she says, they need to find where their cryptography sits and move to the new NIST standards.
What is post-quantum cryptography, and when do you need it?
Lory Thorpe argues organisations should prepare for post-quantum cryptography now, because data stolen today can be decrypted later. They need to find where their cryptography sits and move to the new NIST standards.
Why must healthcare and defence protect data now?
Long-lived data is at risk already. Thorpe explains that attackers can steal encrypted data now and decrypt it once a cryptographically relevant quantum computer exists, so sectors such as healthcare and defence need to protect it today.
Are post-quantum cryptography standards ready to use?
The first post-quantum standards exist and are ready to use. Thorpe says NIST, which began its post-quantum programme in 2016, standardised its first three algorithms at the end of August, and IBM co-developed two of them. She cites Apple's hybrid scheme in iMessage as an early example.
What is crypto agility?
Most organisations do not know where their cryptography is. Thorpe compares the task to being told the nails in a house built over 30 years will one day disintegrate: “Do you know where all the nails are?” Her goal is crypto agility, the ability to swap cryptography out quickly when a threat appears.
What is cryptographic debt?
Every new system built with old cryptography adds to the bill. Thorpe calls this cryptographic debt, and says next-generation telecoms such as 6G should carry post-quantum cryptography from the start rather than be fixed later at greater cost.
Can organisations leave post-quantum security to vendors?
Leaving it to vendors is not a plan. Thorpe warns that vendors act on what their customers demand, and says individuals should check what their banks and platforms are doing. She names HSBC as one of the most proactive companies in this area.
Where Mark and Thorpe disagree
Mark reads out coverage arguing that reports of Shanghai researchers cracking RSA were overblown, and asks whether the danger itself is overblown. Thorpe concedes “Maybe, maybe there is”, but argues that harvest now, decrypt later makes the risk a present one for data that must stay secret for years.
Links from this episode
Questions
What is post-quantum cryptography, and when do you need it?
Lory Thorpe argues organisations should prepare for post-quantum cryptography now, because data stolen today can be decrypted later. They need to find where their cryptography sits and move to the new NIST standards.
Who is Lory Thorpe?
Lory Thorpe works with the IBM Quantum team on post-quantum cryptography after decades in telecoms, from 2G to 5G.
What does harvest now, decrypt later mean?
Thorpe explains that attackers can steal encrypted data now and decrypt it once a cryptographically relevant quantum computer exists, so sectors such as healthcare and defence need to protect it today.
Which encryption does quantum computing threaten most?
She explains why data harvested now can be decrypted later, and why public key cryptography such as RSA is the main target.
Transcript
Jeremy Gilbertson Disruptors and curious minds. Welcome to another episode of thinking on paper. My name's Jeremy. This is Mark. We talk about people building the future rooted into what we're doing today, looking down the road to tomorrow and trying to make sense of it all. There's a lot of stuff coming at you. There's a lot of things on the horizon and we're busy enough. I've got four kids. I'm busy enough just trying to get my kids dressed and out the door. So we're here to help digest this stuff, give some tangible takeaways. Our topic today is amazing in light of our quantum season and what happens when quantum computers figure out how to break stuff, Mark, how are you feeling, man?
Mark Fielding Overwhelmed with stringing these episodes together with an amazing quantum season. And today, today bridges my two interests. So my writing is all about blockchain and crypto quantum season. My curiosity is all about quantum. So this episode will combine those two interests. I'm very excited about it. I think there's a little disclaimer. So we're going to be talking about post quantum cryptography. We, as we've said, we're quantum dangerous, we're crypto dangerous. So we're going to be using some acronyms that we probably don't understand. So bear with us. We'll be trying to explain these different types of cryptography as we go. So if anything, you don't get anything, just leave a comment and we'll get back to you with those answers. But yeah, looking forward to it, Jeremy. Have you done your homework?
Jeremy Gilbertson Yes, I have. I have done my homework. I always do my homework, Mark. Good. Yeah. So post quantum cryptography, before we bring our guest on, literally this is the point in time on the horizon where quantum computers have figured out how to unlock what cryptography does to protect things, kind of in a very basic realm. So we're looking at a point in the future that's either cryptography has to change a little bit or regulations have to do something to make sure things don't happen. I don't know. There's a whole bunch of things that we can explore, but instead of me pontificating, let's bring in the expert, Mark.
Mark Fielding Yeah, let's bring in. Hopefully we'll get to make some real sense of some of the news stories of late as well about cryptography being cracked or not. So yeah, welcome. Our guest is Lory Thorpe from IBM. She's going to be speaking to us about NIST, the National Institute of Standards and Technology, standards for post quantum cryptography and how the industry is collaborating together to make those a reality. How different industries can prepare, what they should be preparing for and lots of other things. So welcome to the show, Lory, and thank you for your time. Appreciate it.
Lory Thorpe Thank you for inviting me. It's a pleasure to be here.
Mark Fielding Excellent. We're in England this week, so we've gone from America, now we're in England. So that's good news for our English fans, Jeremy.
From 2G to 5G, and into quantum
Jeremy Gilbertson That's right. That's right. So quickly, before we jump in, Lory, on the carryover question from the last show, we had D-Wave on quantum computing, one of the earliest players in the quantum computing game. Give us just a touch of background. I looked at your background a little bit, and you've done a lot in the telecom world. Right? So give us a little background about your experience in telecom and how that translated into security and privacy protection and all that good stuff.
Lory Thorpe Sure. So, yeah, so my background is definitely in telecom, many, many decades. I've been working in the telco space since the very early days of mobile. So I've lived through all of the various G's, so from 2G to 3G to 4G to 5G. For many years, I was a solution architect and security was always very much part of the remit. And since I've joined IBM, I've been working in the context of the IBM quantum team, looking at what is the impact both of quantum computing on telco and looking at the potential benefits that quantum can bring to telecommunications, but also then looking at what does post-quantum cryptography mean, how the telcos can protect themselves and continue to guarantee the security, the privacy that everybody expects from the connectivity that we run, that we rely so much on today. And that is sort of one of the things that I've been doing over the last few years.
Mark Fielding 2G, 3G, 4G, 5G. I remember the early days. It must have been a very interesting process to be part of. It was moving so quickly. Is it still moving as quickly now as it was when the progression from 2G to 3G?
Lory Thorpe I think that at the time, and to give you an anecdote, when I started working in telco, my mum was asking me, why would anybody want a phone that they bring with them? You've got a phone at home. Why would you want to bring a phone with you? What's wrong with you? So that gives a little bit of a measure of how much things have actually changed. I think that telco is in probably a challenging place at the moment in terms of its evolution and the business model that has sustained it up until now, which I think now is facing some sort of challenges. But I still think it's a hugely exciting place. And I also think we shouldn't just be thinking of telco as our phones. We need to be thinking of telco as an enabler for digitalization. It's an enabler for digital society, for digital economy.
Lory Thorpe And that is really what makes a lot of this really exciting, because we're not talking about a single sector. We're really talking about a capability that enables everything we do as a society.
Jeremy Gilbertson It literally is the connected tissue for the digital experience, right? So if you don't have that, we can't have these interactions back and forth. We can't collaborate. We can't share data. We can't work with data. We can't pull it in. We can't push it out. Like, yeah, it's really, really good to think about it from that perspective. So let's do the carry over question. I know we're changing, going from telco to health care real quick, but I wanted to do this handoff because we always like stitching these shows together. So Mark, ask that carry over question we had from D-Wave from last week.
What healthcare should be doing now
Mark Fielding Yeah, Murray Thom, head of evangelism at D-Wave Quantum. His question for you was about cybersecurity from a different perspective, and he was particularly interested in the health care industry. And his question was, what advice do you, IBM or the greater community, provide for health care professionals and what should they be doing to prepare short term and long term?
Lory Thorpe So health care is an interesting one, because if we think of the sensitivity of the data that's involved and we think of the long term impact of that, health care is actually one of the sectors that really needs to be thinking about how to protect data, not just against a future quantum computer, but how do we ensure that data isn't harvested now to be then decrypted later? And the recommendation I would have is to start looking now at, first of all, a plan for post-quantum cryptography going forward, but also how to protect some of that long shelf life, important data that sort of people rely on being kept safe.
Harvest now, decrypt later
Mark Fielding I've got a silly question already, Jeremy. That didn't take long, but silly questions are the best. Something you just said then, you said you don't want the data to be harvested now, to be decrypted later. And I hadn't really thought about it in that way. So we'll get into this, and again, apologies for jumping around a little bit. If the cryptography is updated at a later date, if it's changed to encrypt the data, but if they have the data now that's encrypted in a certain way, and the quantum evolves that it can crack this cryptography, but it can't crack what comes later, if they already have this data that's encrypted in this way, obviously the data can't be back encrypted, so they would be able to hack what they have now. And that would be the same across industries, so military as well. Is that right?
Lory Thorpe That is right. And that's why when we talk about post-quantum cryptography, often people think that this is a problem for the future. So a lot of people ask the question of when will a cryptographically relevant quantum computer appear, which in my view is the wrong question, because we don't know for certain, we don't have a definite date. So if we think of the comparison with Y2K, there isn't a corresponding 31st of December 1999 at 11.59, that's your sort of D-day. Here it's a little bit more uncertain, but I think what we need to be doing is thinking about what are the things that we can do now to mitigate something that we know is going to happen, because we are seeing quantum computers that are evolving.
Lory Thorpe We are seeing algorithms that are being optimized, which mean that actually maybe some of these capabilities will actually appear sooner than people expect. And the harvest now decrypt later. Once data is stolen, and even if it's encrypted, so it can't necessarily be decrypted immediately, but once a cryptographically relevant quantum computer is available, then it can be decrypted. And once it's stolen, it's stolen. So there's not a lot that we can do with data that's been stolen, which is why we need to be thinking about that data that has a longer shelf life. We need to be protecting it now. We shouldn't be waiting for a cryptographically relevant quantum computer to be appearing, because then it will be too late.
Mark Fielding Then you need to call the lawyers. Yeah, absolutely.
Getting governments and clients to take it seriously
Jeremy Gilbertson So IBM Talks 2 is relied upon by a lot of large companies for technological advice and foresight and all of that stuff. Herein lies the challenge with some of these large, actually all companies really. So you have an IT organization that isn't always funded necessarily the way that IT organization would want to be funded, right? Because they're getting all these directives and like, hey, I need you to do XYZ. And then the money that comes to them is not always matching what needs to be done. So a lot of things can get pushed off, right? So if there's no immediate threat, then it's kind of like, all right, well, we can wait, we can wait, we can wait. How are you helping bring in the focus now for some of that stuff? Because are you seeing some of the challenges like, oh, we've got plenty of time on this, and how are you dealing with some of those conversations?
Lory Thorpe So there are different aspects to how we're dealing with it. So the first is working very closely with government, with policymakers, with regulators to help them understand and gain awareness of the threat. And we're also working very closely with our clients to ensure that there is an understanding and awareness. I think the awareness piece is really, really critical, because I think once it is understood, then the right decisions can be made based on risk appetite, based on business priorities. So we are aware that, obviously, there are a lot of priorities and a lot of the particularly the security organizations are under a lot of pressure, because they're seeing a lot of threats that are materializing, and some, potentially, they'll be seeing them as more imminent than the quantum threat.
Lory Thorpe But I think raising awareness is a really critical part of the work that we do. I know we'll be talking about this a little bit later, but the other aspect is the industry level engagement. And here IBM has put a lot of effort into working with different industries and basically, again, raising that awareness, but also coalescing around how to act as an industry, because a lot of these capabilities are not going to be, these problems are not going to be solved by one individual company. It will take, as it were, it takes a village. We need to ensure that we have the supply chain lined up. We need to align with regulation and government. And we need to ensure that all of these different elements fall into place.
Lory Thorpe So if we think of standards, if we think of supply chain, if we think of the products that we're expecting, these are all things be somehow aligned and coordinated. And that takes effort. And IBM has taken a very central role in bringing together industries to be able to do that.
Did researchers really crack RSA?
Jeremy Gilbertson Mark, you talk about recently in the last couple of shows, you referenced something that happened with some researchers in Shanghai that cracked some kind of cryptography. Lory, are you familiar with that? And what are your thoughts on what happened there? And what does that mean?
Lory Thorpe Well, I think over the last sort of few years, we've been seeing a lot of announcements, some more accurate than others. I think the paper that you're referring to was later sort of analyzed and sort of disproved. There had been some errors made in the calculation at a certain step of the way. I think at the seventh step, there was an error in the calculation. So the papers basically said, you know, we've cracked RSA. And I think what that shows is that there are a lot of people that are thinking about this. This is something that a lot of people are thinking about and are working on. So the fact that that particular paper was disproved, that doesn't mean that others are working on this. And it doesn't mean that maybe the next time, actually, there won't be an error in the paper. And we will actually have to potentially accelerate when we thought that this was going to be possible.
Mark Fielding I think that's so from what I know. So they cracked a 20 digit or something RSA code. But in reality, RSA is like 2000 bit cryptography. And obviously, it's exponentially higher. So it's nowhere near that. But that's a nice segue, maybe. So I'm from crypto and Bitcoin uses elliptic curve cryptography or something. There are different types of cryptography. Maybe we could just for more advanced or for our beginners as well, maybe outline three or four of the most commonly used cryptographies, including that one of Bitcoin and the RSA, maybe a little bit in how they differ and how susceptible they are to attack or being hacked first and then move on to the kind of standards that NIST are trying to build around those.
Which cryptography the quantum threat breaks
Lory Thorpe Yeah, well, I mean, I think the first thing we can start with is what is what is sort of majorly affected by the quantum threat is public key cryptography. So there has been some work done on symmetric key cryptography. We believe it's less sensitive to the quantum threat, even though there will need to be some work done to ensure that the key lengths are sort of fit for purpose, which, you know, even now, in some cases, you know, we've we've sort of left cryptography, left sort of old cryptography in there and haven't necessarily been being actively managing that maybe as well as we could have. When we talk about different algorithms that will be affected. When we look at public key cryptography, ultimately think of RSA, think of elliptic elliptic curve cryptography, think of Diffie-Hellman.
Lory Thorpe These are some examples of the cryptography that will be affected by a cryptographically relevant quantum computer. These are the algorithms, but obviously then the implementation of the algorithms, not all algorithms are created equal and not all implementations are created equal. And that is also why we can't just give a single D-day or Q-day, as it were, around when cryptography will be will be compromised. I think one thing maybe to think about is that obviously when when the first cryptographically relevant quantum computers will become available, to be able to craft cryptography, it will be potentially quite an expensive and quite a difficult thing to do.
Lory Thorpe Then as time goes on and the capabilities sort of become more democratized, then it will become easier and cheaper to craft cryptography, which may also determine which are going to be the targets of that cryptography, if that makes sense.
Jeremy Gilbertson Yeah, I don't remember who was on the show, Mark, that kind of put this thought in my brain, but it's kind of more about, instead of taking 10 years and $100 million to break cryptography, maybe it'll take seven and like $50 million and maybe five and $30 million, right? So like the appeal to try it might start to be a little bit more palatable for the malcontents out there, maybe.
Mark Fielding Well, that's their incentives. I think that was Equal1 that put that, Jason Lynch, I think he put that in your...
What NIST standardised, and why it took years
Jeremy Gilbertson Well done, well done. Okay, well, let's move on to the NIST work that you've been doing and maybe give us a high level detail about some of the outputs that you're starting to see as it would relate to a chief technology officer at an organization.
Lory Thorpe Okay, so I guess the first thing is NIST stands for the National Institute of Standards and Technology. NIST, a few years ago, back in 2016, they initiated a program which is the post-quantum cryptography program. And the intention of that program is basically to standardize algorithms that are resistant, not just to classical computers, but also to quantum computers. And so that process, it's very much a global process, and it has involved a lot of sort of entities from all over the world. This includes sort of the entire sort of global cryptography community, if you like. And what they did is they put out a call for proposals for algorithms that they could standardize. And when they first started this process, I think initially there were something along the lines of around 80 algorithms that were submitted.
Lory Thorpe Now, that process, it lasted for many years because what NIST has done, they've done a fantastic job of basically looking at the security of the algorithms, but also looking at things like the performance of the algorithms in an attempt to standardize the best ones. What has happened recently, which is very exciting for us, is we've reached the first milestone where they've standardized the first three post-quantum cryptography algorithms. And this happened at the end of August. Those first three algorithms are, we're happy to say that IBM was co-developer in two of those three algorithms. So they selected these algorithms where the draft standards were issued. Now we have the standards.
Lory Thorpe And what that means is that effectively, we've got a first set of standards that we can start to implement into products, into protocols, to be able to start to make our systems quantum safe. So it's a huge milestone for us. Very, very excited because obviously this is, if you like, it's a starting point for the implementation.
Can you use the new algorithms today?
Mark Fielding And can those three algorithms be used today in, if somebody made another WhatsApp, for example, it could be incorporated that as a means of encryption. If somebody made a new meme coin on Solana, they could use that encryption today to make, and would that be quantum resistant to an indefinite future or a hundred years, a thousand years or forever?
Lory Thorpe Well, I think, I mean, security is, saying that something is indefinitely secure is always challenging. You know, there's always, I think it's tempting fate in some ways. Everything is secure until it isn't. So we could argue that the algorithms have gone through very, very rigorous testing and we have absolute confidence that they are secure. It doesn't mean that they'll be indefinitely secure. And maybe later we can talk a little bit about crypto agility because I think that's a really important part of the whole post-quantum cryptography discussion. But fundamentally, yes, the answer to your question is, yes, we can start to implement them. In fact, in the case of IBM, we started a couple of years ago, we implemented what then became one of the standardized algorithms.
Lory Thorpe We started to implement in our mainframe, in our Z16 platform. And this was really a great, first of all, a great learning curve because understanding how to implement the algorithms is sort of going to be part of that journey towards a quantum safe architecture. But yes, I mean, the short answer is definitely we should be starting to implement them. And even if we look at what, for example, Apple have done recently, so they have implemented a hybrid scheme in the iMessaging solution and they are using the post-quantum cryptography algorithms that have then since been standardized.
Jeremy Gilbertson Okay. Now, Mark, no Apple rants today. We're going to have to hold you back on the Apple rants. But you, so Lory, you had me.
Mark Fielding No, that's a plus one, I think, for Apple, that kind of like they've gained a bit more, a bit of their respect back.
Crypto agility, and knowing where your cryptography is
Jeremy Gilbertson Oh, okay, okay. Lory, you had my curiosity peaked at crypto agility. Can you explain that? Yeah, let's not wait.
Lory Thorpe Okay, so crypto agility, it's, so we've been very, very fortunate over sort of the last few years, the last few decades. The cryptography that we've been using has actually held up very, very well. And, you know, with the, obviously with the few exceptions there, we have done some cryptography migration, but actually there's a lot of cryptography out there that has held up for decades. And that is really good news. But what it's also done is it's put us in a situation where potentially we may have become a little bit, what's the word, complacent in how we manage cryptography.
Lory Thorpe So one of the big challenges that we have today, and we're seeing this across different industries, one of the big challenges that we have is that we don't necessarily know where and how cryptography is being used in the systems that we use today. And this is doubly true in the sectors where there's a lot of legacy. So if we think of financial services, or we think of telco, where potentially you have multiple generations of technology that are being sort of built one on top of the other, knowing where that cryptography is actually being used and how it's being used, that is not at all obvious. Now, what does that mean?
Lory Thorpe It means that if something happens where the cryptography needs to be changed, like as is the case for the quantum threat, the first thing we need to do is understand where we're using the cryptography, and understand how relevant it is in terms of what it's protecting, what would be the business impact of sort of if that cryptography were to fail. But also it means that we're not really in a position to very quickly swap out cryptography. And going forward, becoming quantum safe is a little bit of a catalyst to maybe managing cryptography better going forward, and ensuring that we're building security in a way that is able to, in an agile way, respond to threats going forward, whether it's a quantum threat or whether it's a different threat. And that's why we're really building that crypto agility message as part of the quantum safe journey.
Is the threat overblown?
Mark Fielding Very interesting. Jeremy, I don't know if you saw, but I know it's kind of crypto agility, but Vitalik Buterin was talking about, with Ethereum, that one of their post-quantum solutions is to hard fork Ethereum, and then I guess they would input data, reprogram it with this new algorithm to make it quantum secure. Do you mind if I push back? So, and although this question isn't really very useful now, because what you said about kind of stealing the data and decrypting later makes it a little bit less valid. I think that's, for me, that's the big takeaway is that it doesn't actually matter how good the cryptography gets in the future. If you steal the data now, you can decrypt it. And obviously some of the information that you steal, it will be diluted over time, but a lot of it won't be.
Mark Fielding If it's however many years in the future, if you get hold of it, if the wrong people, if the nefarious parties get hold of it, then it could be detrimental to everything. I want to just read, it's from a, I'm not sure who it's from. It says, current estimates are that quantum cracking of 1048 bit RSA key would require a computer with 20 million qubits running in superposition for about eight hours, okay. The coverage of the September, so those Shanghai researchers, is essentially overblown because symmetric encryption, unlike RSA and other asymmetric siblings, is widely believed to be safe from quantum computing. As long as bit sizes are sufficient, PQC post-quantum cryptography experts are confident that AES256 will resist all known quantum attacks. Is there a threat of overblowing the threat or are we not? Are we overblowing the potential danger of this?
Lory Thorpe I mean, that is kind of a million dollar question, but I guess maybe just to the point that you made around the AES256. So there are two, I mentioned earlier around symmetric and asymmetric cryptography. So the main risk that we see around a cryptographically relevant quantum computer is down to Shor's algorithm, and that's why it affects asymmetric cryptography mainly. That doesn't mean that symmetric cryptography will be completely unaffected, and I think that you're right. I think we need to be looking at key length and we need to be making sure that there is sort of sufficient, you know, that we've got sufficiently long key length implemented. Now, does that mean it's 100% sort of, you know, we can definitely say 100% that it won't be affected?
Lory Thorpe Again, I think that's always a dangerous thing to be doing when we talk about security. So I think a lot of these things, we need to be looking at what is really the art of the possible and how that evolves and what are the things that are actually causing that risk to materialize. So on one side, obviously, you've got the evolution of the performance of quantum computers. You'll have things like error mitigation and error correction that is evolving and quantum computers are becoming more and more powerful. And you've also got ultimately algorithms that are being optimized and that are becoming more effective and more efficient at doing certain things.
Lory Thorpe And I think the combination of all of these things happening, you don't know what it is that is actually going to cause, for example, something to all of a sudden become possible. I don't know if that has really answered your question. Is there a risk of this being overblown? Maybe, maybe there is. But on the other hand, we need to be thinking about, are we ready to take that risk? And then going back to the harvest now, decrypt later, I mean, that risk is here now. So depending on what data you're sitting on, what is the data that you're protecting? If the data has a short shelf life, then you might say, well, actually, I don't see it as a huge risk. But if the data that you're protecting is data that is going to be relevant in 10 years time.
Lory Thorpe So if we think of health care, if we think of defense, if we think of some of the areas where data is really, these are the crown jewels. It depends how gung ho you want to be with your security and with the data that you've been tasked to protect.
Why starting early is the whole game
Mark Fielding Yeah, yeah, I think that's the key. I was going to ask Jeremy to paint his worst case scenario. And if you're taking the risk of that coming to fruition, then yeah, why take the risk? Sorry, go ahead. No, please.
Lory Thorpe No, I was just going to say that the other thing maybe to consider is that the implementation of post-quantum cryptography, this isn't something that we are going to do next week. This is something that takes a significant amount of effort and time. So one of the things that we are doing is looking at how we can make it more efficient, more effective and more secure. And really, the answer to that is the sooner you start in that process, the better you'll be able to manage the journey.
Lory Thorpe So if you think of how you can, for example, combine your post-quantum cryptography transformation with other technology transformations that are going on in parallel anyway, that is sort of one of the areas where actually you can do it a lot more efficiently if you start early rather than waiting and then having to sort of do it either in a rush or sort of go back and fix things that you're building today. So one of the things you don't want to do is to build more cryptographic debt.
Cryptographic debt
Jeremy Gilbertson What do you mean by cryptographic debt? I think I know where you're going, but what do you mean by that?
Lory Thorpe So if I'm implementing a new widget and I have the opportunity to put, for example, to secure it against a quantum risk, so to implement post-quantum cryptography and secure it against a post-quantum risk, that means I won't be increasing the cryptographic debt. If I'm implementing the new widget and I decide that I'm just going to put old cryptography in there, classical cryptography in there, that means that at some point I need to go back and change that cryptography. So there is a case for particularly if we think, you know, I'll go back to telco. If we think of the next generation of telco capabilities, I want to be implementing post-quantum cryptography from the start.
Lory Thorpe So one thing that I don't want to do in the case of, for example, 5G evolution or 6G, is to implement classic cryptography that I then need to go back and change. Because that will, effectively, that will cost a lot more as well as sort of having an impact in terms of effort. So where possible, we want to be building post-quantum cryptography into any new capabilities that we implement as soon as possible.
Jeremy Gilbertson Yeah, it's like Mark, you go into the pub and just keep putting everyone's tab on your credit card and then dealing with it down the road.
Mark Fielding Yeah, it's always going to come back and bite you.
Jeremy Gilbertson That's right. No, that makes a lot of sense. I like that. And, you know, where are my heads going with a lot of this? And I've got a fun thought experiment that we could run into if you're up for it. And then maybe we can deal with our takeaway question. But what I gathered from listening to you is the idea of the cryptographic agility kind of thing is really interesting because it's no longer a set it and forget it kind of thing. It's no longer, hey, did you click the security button? Are you doing your thing? Yeah, it's just kind of good. It's more of, hey, we put something in today, but we got to monitor it. And maybe there are some things that we could do on a monthly basis to increase that, to add code to it, to do something a little bit better with that.
Jeremy Gilbertson So I thought that was an interesting takeaway because a lot of people are just like, yeah, my stuff's encrypted. Well, let's unpack that a little bit and see. So that's really cool. All right. My thought experiment for you and, you know, no is a perfectly acceptable response to this, but let's imagine during this interview, you get a call, like a WhatsApp just pops up and it goes, hey, yeah, Lory, so here's what's happening right now. It's like your NIST, your buddies from NIST. So it looks like they figured out a date. It's going to be November 1st, 2031. Quantum is going to be all cryptography. That's where we are. We got to get to work. How are we sitting today? And, you know, what would the message be to everybody if we now have like the asteroid date, you know, up and running as November 1st, 2031?
Lory Thorpe Well, to be honest, I think the answer, regardless of whether we have a date or not, I think the answer would be the same, maybe with slightly different senses of urgency, but I think the answer would be the same. I mean, the recommendation would be start to get a plan together. Start to do an analysis of where you're using cryptography, how you're using cryptography. And I mentioned this earlier. This is sort of far from obvious because sometimes people don't really think about cryptography too much. But one of the analogies that I like is, you know, imagine you've built a house and imagine you've built that house over the last 30 years and the house is there, it's standing, everything is working as it should.
Lory Thorpe And someone comes and says, you know what, the nails that you've used for that house, they're going to disintegrate at some point in the future. Do you know where all the nails are? And cryptography is a little bit like that. Because actually even understanding where and how you're using cryptography in a complex enterprise, that is a really good starting point. And it takes a lot of time to be able to do that. It takes a lot of time. And there are a lot of challenges in doing that. This is why one of the things that IBM has been doing is trying to automate some of that process to be able to get that information. We've also been standardizing some of the, sort of how the information is actually ingested.
Lory Thorpe So some of the work that we've been doing with the cryptography bill of material really is to try and get to a point where it's easier to gain that information than it is today. Because today, we don't have a standard way where our vendors are saying, okay, you bought this, here is how this is how we're using the cryptography. So there are a lot of challenges to be able to get an understanding of where cryptography is being used, which means it can also be challenging to understand what the impact of that cryptography being broken is going to be for the organization. So before you can even start to go and start to fix things, you need to really understand where do the risks lie? What is the actual threat? What are the things that you need to start working on as a priority?
Lory Thorpe And what are maybe the things that are lower priority? So that would be my recommendation would be to start getting on top of the cryptography estate, start looking at what are the things that you want to protect as a priority. So the first one would be harvest now, decrypt later, what are the things, what are the the mitigation or remediation solutions that I can put in place today to ensure that my data is not harvested, for example, or sensitive data with long shelf life is not harvested. Some things may not matter. So just because something is vulnerable doesn't mean we need to fix it. We want to be fixing the things that are really important.
Jeremy Gilbertson Yeah, there's a prioritization of which nails are going to disappear faster. Are those nails at a joint on the house? That's a wonderful analogy.
Where cryptography already sits in your day
Mark Fielding On those nails, so not to give people nightmares who are listening to this, but where is cryptography being used in our daily life that people who are not involved in this would be surprised about it? Because people might be listening to this, you have no impact, no crypto experience, no quantum experience going on. It doesn't affect me. And where?
Lory Thorpe Well, I think, I mean, ultimately, I think cryptography affects us all, whether we realize it or not. So, you know, who, you know, who's used their phone today? Who's gone and done some online banking or who has sent an email or logged into, logged into, you know, made an Amazon purchase? So these are all areas that ultimately cryptography is involved. So is there anything in our digital life that isn't anything in our digital life that isn't? So if your digital life is based on cryptography?
Is it somebody else's problem?
Jeremy Gilbertson Well, here's the, here's the philosophy too, that's happening right now is that most people rely on these platforms, these apps, these organizations to, oh, they've got it all encrypted. It says right here, like all my stuff is encrypted on WhatsApp. It's great. But, you know, then what if that team is not, you know, staying on top of where the disappearing nails are, and then it starts to trickle down to people with it. Right. So I think, yeah, there's an, there's almost like an offloading of risk on the individual level, I think, to the platforms, wouldn't you say?
Lory Thorpe I think, I mean, I think there is a risk that people just think that it's somebody else's problem. And, you know, and, and even with some enterprises, I think, you know, one solution could be, well, my vendors are just going to take care of it. So I can just sit back and relax, which could be an approach. I'm not sure that it would be an approach that I would be comfortable with, because ultimately, we, you know, our vendor, the vendors will, will do what, what the the companies that are buying their services require. And if that require if that market demand isn't coming in loud and clear, they might be sort of, they, they might think that it's okay to, to postpone that.
Lory Thorpe So I think there's, there's, we all have a responsibility to make sure that the things that we're responsible for that we're accountable for that, that these things are really happening. And then if the vendors are doing what they what, what we expect them to do, then that's great. But I don't think it's something where we can just sit back, relax, and rely on somebody else to to take ownership of that problem. As individuals, one of the things that that certainly I'll be doing is looking at the platforms that I'm using and looking at what they're doing to protect my data against a quantum threat or any other threat. I mean, quantum threat is, is the example we're using here, but that would equally apply to any other threat that that might be coming.
Lory Thorpe So if we think of our online banking, I would want to know that HSBC is doing the right things. And, and, you know, and I've used that as an example. And actually, HSBC are probably one of the most proactive, proactive companies in this space. But I think we all have a responsibility to take ownership for this as a risk.
What AI changes about all of this
Mark Fielding This is just one last question before I hand it over to Jeremy for the closing thoughts. And this might be a very short answer, it might be a very long answer. But you talk about handing over responsibility to somebody else. And so we're in the AI era of handing over all our tasks to AI. How, if at all, has the explosion, the acceleration of AI affected how NIST and IBM and of your collaborators are thinking about this? Are you using AI to help you? Are you using, is AI hindering you? Has AI made it more urgent to get this done? Like, how is AI fitting into the NIST, quantum cryptography world?
Lory Thorpe I mean, that's a good question. Ultimately, AI is a very powerful technology, as is quantum. And, you know, in the words of Spider-Man, with great power comes great responsibility. I mean, these are all technologies that are hugely powerful, have huge amount of potential to really change things for good. But they also have the ability to be used for nefarious purposes. And I think that's something that we need to accept, that that is the case for all powerful technologies. So AI, in particular, when we talk about the PQC transformation journey, we're using AI as part of the work we do around automation and around being able to interpret data and around being able to make that transformation more efficient, more effective, more accurate. So that's the part where, you know, we're using AI to help.
Lory Thorpe And we're using AI to make things, make things better. But equally, I think we can look at what does AI mean for security. And there are people that will be using these technologies for nefarious purposes. And we need to be prepared for that. And we need to take a responsible view of how these technologies can be used. And this is sort of, this also goes back, so I'm part of the IBM quantum team, our mission statement is on one side to bring useful quantum computing to the world, on the other, it's to make the world quantum safe. So I think we, you know, we need to be looking at all of these technologies under this sort of dual hat, if you like, how do we leverage them for the good that they can bring, but also how do we protect ourselves against the potential damage that they can do? I don't know if that answers the question.
Mark Fielding Yeah, I noticed maybe even a tinge of worry in your voice when I said AI as if there's something that it's, it's so incredibly powerful that yes, we need to move quicker than we were.
Lory Thorpe Well, I, it's, I guess it's it's down to, you know, all technology can be used for good and not so good. And I think AI is, is one of the, it is one of those technologies that does, you know, does create some concern. So, because ultimately, it is moving so quickly. And because sometimes, you know, we tend to take a little bit of a, you know, the, the power of the technology is not necessarily understood. And that's where things could potentially go, go wrong. And, you know, and we've seen, and I think AI is a very good example of that. So that's not to say that AI is bad, because I think AI is, again, it's, you know, hugely powerful, and can be hugely beneficial.
Lory Thorpe But we do need to be careful and make sure that, you know, we look at, we look at both sides of the equation before sort of jumping into, into some of these things. And that's why I think that the quantum threat aspect, that's something that actually a lot of really clever people have been looking at for a long, long time. That's why we're in the position we're in where we've actually got standards today, that can help us to move forward towards a quantum safe, quantum safe situation. And that's why sometimes when people say, well, can I wait? Well, yeah, of course, you can wait. Should you wait is another question.
Jeremy Gilbertson 100% agree. So disruptors and curious minds, some some takeaways that are swirling around in my brain. You know, as you're a user of these technologies, using apps and platforms, you know, as Lory said, you know, start to consider what they're doing in a post quantum world or to prepare for a post quantum world as an individual, and businesses, IT department, CTOs, listen, you can either get started on this, I think, or you can start deleting all of your data and printing, you know, hard copies of it, stuffing it in a, you know, lockable shelf somewhere, and then crossing your fingers. So, yeah, so that's the thinking on paper, that's right, we're going back to analog, you know, we're all going to have a spot in this. So, Lory, this has been a fantastic conversation.
Jeremy Gilbertson I've really enjoyed your insights and your work is going to lead to some very helpful things for people and companies down the road. Let's finish with our carryover question. So next week, you know, speaking of AI, next week, we're talking to a studio, some executives from some large creation houses that are using AI to create animation. What would be your question for them, you know, related to anything? It doesn't have to relate to security or cryptography or anything, but what would you leave for them as a handoff?
Lory Thorpe Well, I'm going to continue with the theme that we've been discussing here today, and I would ask them, you know, I think the entertainment business is a really interesting one because of all of the sort of the copyright implications. And so, you know, what are they, what are they, have they started thinking about their transition to post-quantum cryptography?
Jeremy Gilbertson I like it. I like it. Yeah, they're all about the kind of new IP, so the copyright question is very pertinent for what we'll be speaking about. Mark, tell them about the book club, and we'll get on our way here.
Mark Fielding Do you know what? For the first time, I think, in 120, however many episodes, I don't have a book next to me because we've got a book club. There we go, where we read books that have stood the test of time, books that will change your mind. We are the rebellion to the 15-minute book summary. We are the rebellion to that. We're going to read 150 books in a week. We're going to read one book very, very thoroughly together and then bounce our ideas around to try to understand, to read between the lines, to make sense of books. And we're reading, as Jeremy showed you, or if you're just listening to this, Quantum Supremacy by Michio Kaku, who still hasn't agreed to come on the show, by the way.
Jeremy Gilbertson Still working on that one. Still working on it. You got any connects? IBM, you have any connects to him?
Lory Thorpe We'll see what we can do.
Mark Fielding And the next book will be Nexus by Noah Harari.
Jeremy Gilbertson You and Noah Harari. And I don't think he'll be coming on the show either unless you know somebody there. But yeah, join the book club, ThinkOnPaper.xyz. Like, subscribe. And Lory, thanks for joining again. We'll put out some notes. And if you have anything, any links or anything you want to send Mark, we have a great post-show wrap-up that we do. But thanks for joining. Audience, be curious, stay disruptive, keep thinking on paper.
Lory Thorpe Thank you so much for having me on the show. Thank you for




