Raghda Ibraheem: This week, outages hit Chase and other US banks, Apple Pay and Charles Schwab’s website, leaving customers unable to reach their own money. Companies point to software bugs and internal updates, but the run of failures raises a bigger question: how fragile are the digital systems behind our money, and what should the fast-digitizing MENA region learn from it?
Joining me is Roxana Nasoi, Community BD Lead at Logos and core contributor at the Institute of Free Technology. Roxana, welcome to Capital Markets. Now, this week brought a wave of outages across major systems, including Chase and other US banks, Apple Pay, Charles Schwab, Ikea as well, the BBC and a lot of websites. So what does this run of failures tell us about how fragile the digital systems behind money really are?
Roxana Nasoi: Thank you for having me. Yeah, I guess it’s probably something that we’ve seen across the years, especially, you mentioned Equifax. In 2017, they had one of the largest US cyber attacks on identity theft. It affected about 149 million US citizens, 15 million British citizens, and 19,000 Canadian citizens.
It was essentially a vulnerability patch that was left hanging that the attackers could exploit, which led to a theft of identity information: Social Security numbers, addresses, credit card information, and so on. And it had a long-term effect. And so whenever I hear about cyber attacks, my main concern is the long-term effects.
And I feel like we have not learned anything in the past ten years or so, because here we are today. The same kind of fragile system introduced by central points of failure, introduced by centralization, and by these kinds of systems that lowered the cost and lowered the resources necessary, brought by big tech companies.
Raghda Ibraheem: Yeah. Same problem, same mistakes, I think. And companies, as you are seeing now, have blamed software bugs and internal updates, and some have raised the possibility of cyber attacks. Now, why is it so hard in the first hours to tell the difference between a technical failure and an attack?
Roxana Nasoi: Mostly because they look quite similar. So in the early hours, attackers a lot of the time want to mimic operational failures. So you may end up with an error accessing certain functionalities, or even, you know, an admin reset email. But the reality is this is done by design. And again, this shows that the failure is structural.
So we need to, we’re in a stage right now where we need to think of alternatives and think of building systems that are designed with privacy in mind, and also designed to minimize failure at what we call the lifecycle of a transaction. So, for example, what we do at Logos, we have an infrastructure design philosophy called PriFi.
Because from private finance we’re looking at seven steps, the seven points that occur throughout the lifecycle of a transaction, and that is from discovery to diligence to negotiation to contracting to settlement, [unclear: “ordering”] and enforcing. And so each of these steps comes with sets of risks and sets of data associated with it.
So minimizing the security risks around this requires dedicated infrastructure, which you don’t see. This kind of philosophy is not embraced currently by the financial system, but also systems that handle records, personal data and so on.
Raghda Ibraheem: Yeah.
Roxana, you work on digital identity and privacy. Now, when these systems go down, what happens to people’s access to their own money and their own data?
Roxana Nasoi: Well, as expected, if you can’t control, if you can’t participate in the governance of an infrastructure, you are essentially at the mercy of the intermediaries. So a lot of the time your access will be revoked. You won’t be able to access your medical records, you won’t be able to access your financial records.
So it’s pretty brutal. And again, the only solution is having a way to mitigate risks. A lot of the time we place a lot of faith in third-party ventures and infrastructure that is centralized, that has central points of failure. And unfortunately, at the end of the day, the consumer is at risk.
Raghda Ibraheem: Yeah, definitely. Now, last thing here, Roxana: does AI make these systems more vulnerable, as some experts argue today, or is it too early to say or assume?
Roxana Nasoi: I would say we’re seeing already the capabilities of AI, right? AI doesn’t have just pattern recognition capabilities. It has pattern creation abilities. So it will be able to find vulnerabilities faster. It will be able to design attacks faster than any human being.
And so the problem is that we rely on very old code that probably has a ton of bugs, a ton of security vulnerabilities that have not been patched. So again, it’s a matter of not saying that AI is bad, but building solutions, building parallel infrastructure options that provide these institutions, these organizations, a way to switch whenever an attack on their traditional infrastructure is happening. And minimizing the risk, of course, from these mistakes.
Raghda Ibraheem: For sure. Yes. Thank you so much, Roxana, for being with us on the show today on Capital Markets.
Roxana Nasoi: Thank you.