The X-Factor Weekend Roundup: When Financial Systems Are Put to the Test | November 16, 2026
Crypto Hipster00:50:0145.8 MB

The X-Factor Weekend Roundup: When Financial Systems Are Put to the Test | November 16, 2026

Five stories. Five institutions facing questions about trust, protection, and accountability.

In this week's edition of The X-Factor Weekend Roundup with the Crypto Hipster, Jamil Hasan examines five developments shaping the relationship between technology, finance, and institutional responsibility.

This week's stories:

  1. Microsoft 365 Outage: What a major service disruption reveals about operational resilience and our dependence on technology infrastructure.

  2. Adam Back and Bitcoin Custody: Why simpler security architecture may offer advantages over increasingly complex custody arrangements.

  3. OpenAI and AI-Powered Payments: As ChatGPT moves toward payment capabilities, who controls authorization, transactions, and accountability?

  4. Congress and Prediction Markets: House Oversight scrutiny raises questions about government officials, privileged information, and financial market integrity.

  5. Project Meridian FX APAC: Central banks explore how traditional payment infrastructure and distributed ledger technology can work together to improve cross-border settlement.

These developments raise a common question: When technology changes how institutions operate, who remains responsible for the consequences?

No hype. No price predictions. No partisan politics.

Just evidence, institutional accountability, and the questions worth asking.

The X-Factor with the Crypto Hipster
Monday Weekend Roundup | November 16, 2026

Hosted by Jamil Hasan
CHP Publishing | Crypto Hipster

[00:00:03] [SPEAKER_01] Welcome back to The X-Factor with The Crypto Hipster. This is the Monday weekend roundup for week four for November 16, 2026. It's five stories. We have five stories today, five developments worth examining, and five opportunities to look beyond the headlines and ask what is actually happening.

[00:00:30] [SPEAKER_01] Again, my name is Jamil Hasan. This is the X-Factor Weekly Roundup. We are here three times a week. We are here on Monday mornings to bring you the weekend roundup. We are here on Wednesdays to bring you the prediction ledger where we examine the people's Bitcoin's predictions.

[00:00:51] [SPEAKER_01] And then Friday, we do a full weekend week wrap up called The X-Factor with The Crypto Hipster, which is our flagship news story, which I brought back after having off a year and a half after doing it for a year and a half back in 2022 to 2024. So this weekend we looked at a major Microsoft outage that disrupted communications around the world.

[00:01:22] [SPEAKER_01] We looked at Adam Back's argument for simpler Bitcoin custody security. And we looked at open AI's move into AI-powered payments. We examined a congressional investigation into prediction market trading by government officials. We did that Friday. We're doing it again today. Different story, different angle. You know, it's important.

[00:01:51] [SPEAKER_01] And then a central bank experiment connecting traditional financial infrastructure with tokenized settlement systems. That one you may not know about. These are different stories involving different technologies, institutions and responsibilities. And that's exactly how we're going to approach this. We're not here to turn every technological development to a resolution.

[00:02:17] [SPEAKER_01] And we're every failure into a catastrophe. Or every announcement into evidence of progress or drama or entertainment. Sorry, not everything is entertainment. We're here to examine the facts, challenge the claims and consider what the developments actually mean.

[00:02:46] [SPEAKER_00] So, let's begin. Segment one.

[00:02:57] [SPEAKER_01] Microsoft 365 went down on Thursday. We're going to look at what happens when critical infrastructure fails. Because on Saturday, two days afterwards, Microsoft confirmed that the services affected by the global outage had been restored. Yay.

[00:03:18] [SPEAKER_01] The disruption began on Thursday, continued Friday, affected the outlook, exchange online, portions of Microsoft Teams, and other services connected to the Microsoft's cloud infrastructure. The outage affected businesses, government agencies, and other organizations across multiple countries. It also created problems for third-party services that depend on Microsoft systems.

[00:03:46] [SPEAKER_01] OpenAI reported disruptions involving its Microsoft 365 connector and chat GPT. In fact, I was working on a project in chat GPT and chat GPT, he couldn't recall anything about the conversations that I've had with it over the past several months regarding the project. Well, I was able to restore it by getting a JSON code. But, a lot of people don't know what a JSON code is.

[00:04:16] [SPEAKER_01] Anthropic experienced similar problems involving Claude's connection to Microsoft services. This was a big deal, other than just email. According to Microsoft explanation, this was not a cyber attack. And it was not an AI agent going rogue or making its own decisions on behalf of humanity. No. And it was not a hostile actor penetrating the network. So, what was it?

[00:04:47] [SPEAKER_01] Microsoft attributed the disruption to a technical problem involving an earlier network capacity upgrade that became apparent following the subsequent network change. We used to have that at AIG. We would take care of it right away. Within a couple hours, everything was restored. But this is much bigger than what AIG was. So, there are several things here worth examining.

[00:05:13] [SPEAKER_01] First, we need to distinguish between a cyber security incident and an operational failure. We tend to discuss, you know, infrastructure security in terms of attackers. Especially in North Korea. And we should shut down North Korea, by the way. But that's a different segment on a different show. We talk about ransomware, credential theft, malware, unauthorized access. And those all are legitimate concerns.

[00:05:40] [SPEAKER_01] But systems can become unavailable without anyone attacking them. A configuration change can cause a failure. A capacity problem can spread through the interconnected services. A recovery procedure can take longer than expected. The result for the customer can be very similar. You cannot access your email, which most people could not. You cannot retrieve your information, which I could not. And other companies could not.

[00:06:10] [SPEAKER_01] And you cannot perform ordinary business operations. It's a problem. And depending on how your organization works, you may not have an alternative available. That's the first issue. The second issue is dependency. Microsoft 365 is not simply an email product anymore. For many organizations, it is the underlying critical infrastructure.

[00:06:38] [SPEAKER_01] Through which employees communicate, coordinate, schedule, exchange documents, and interact with other applications and with their clients. And as more AI products integrate with these systems, the dependencies become more complicated. Think about what happened here. Microsoft infrastructure problem affected Microsoft customers and many other customers outside of Microsoft.

[00:07:06] [SPEAKER_01] It affected outside AI services that relied on Microsoft's infrastructure to perform tasks. Those AI companies did not necessarily cause the failure, yet their customers experience the consequences. This is something that deserves much greater attention. Going forward, as software becomes more interconnected.

[00:07:28] [SPEAKER_01] We talk about AI making business decisions more productive and we talk about agents retrieving information, coordinating tasks, and automating workflows. But what happens when a system supplying that information becomes unavailable? It's supplying anything. Does the agent recognize that the information is missing? Chat work? I don't remember anything we talked about.

[00:07:57] [SPEAKER_01] That's a problem. Does it stop? Does it notify the user? It didn't, to me. Does it attempt to continue using incomplete information? Yes. It did. Those are different questions and they require different safeguards. I also want to be very careful about drawing the wrong conclusion here because it's important to not draw the wrong conclusion. This outage does not establish that cloud computing is fundamentally unreliable.

[00:08:25] [SPEAKER_01] There are a lot of great innovations in cloud computing area. It also does not prove that every organization should abandon Microsoft. You should not. What's better? Ask yourself that, right? It certainly does not establish that decentralization would have prevented the problem. A lot of your decentralized protocols rely on Microsoft, by the way, or Amazon. You'd be surprised.

[00:08:57] [SPEAKER_01] Decentralized systems have their own vulnerabilities, coordination challenges, and failure modes. The relevant question is whether organizations understand the dependencies they have created. If email becomes unavailable for a day, what happens? I pick up a phone? Text? You're not going to send important confidential information via text, are you? Text?

[00:09:23] [SPEAKER_01] If an AI-connected service loses access to corporate data, what happens when you need that corporate data? If employees cannot reach the tools they normally use, what processes remain available to them so they can continue with business continuity? Business continuity is not something that should be addressed only after an outage.

[00:09:47] [SPEAKER_01] There's also an accountability challenge here because Microsoft has said the incident was resolved and promised a post-incident report. That's important, right? But a meaningful report should do more than identify the technical trigger. It should explain why the issue wasn't detected before it affected customers.

[00:10:08] [SPEAKER_01] What safeguards failed to contain it and how the recovery process worked and what changes will reduce the likelihood or impact of a similar incident. Not because every outage can be prevented, it definitely can. You're going to have it. It's going to happen. Because organizations that provide critical services have a responsibility to learn from failure. It's called making an amends.

[00:10:37] [SPEAKER_01] And the customers need enough information to make informed decisions about resilience. Only the resilient organizations survive the long haul. The lesson here is not that technology failed and therefore technology cannot be trusted. We can still trust technology. The lesson is that availability, recovery, and transparency are all part of the value of a service.

[00:11:02] [SPEAKER_01] A system does not become dependable simply because millions of people use it. It becomes dependable through engineering, through operational discipline, through redundancy, and through demonstrated recovery and doing the right thing.

[00:11:27] [SPEAKER_01] And even then, organizations need to prepare for the possibility that it will become unavailable. It's an ordinary operational change that produces extraordinary consequences.

[00:11:46] [SPEAKER_00] And that's where we'll leave Microsoft.

[00:11:57] [SPEAKER_01] Segment two, Adam Back, Blockstream, and the case for simpler Bitcoin security. So the second story takes us into Bitcoin custody. On November 13, Blockstream announced a new institutional digital asset custody solution developed with French technology company Fails.

[00:12:20] [SPEAKER_01] The system combines Fails hardware security models with Blockstream secure multi-party computation technology. According to Blockstream, the architecture allows institutions to authorize transactions while keeping sensitive cryptographic signing operations inside dedicated security hardware. Adam Back, Blockstream's co-founder and chief executive used the announcement to make a broader argument about security.

[00:12:52] [SPEAKER_01] His position is that increasingly complicated custody systems can create vulnerabilities that become difficult to understand audit and control. He specifically raised concerns about the internet-connected infrastructure and complex multi-party computation protocols. And he suggested that advances in AI could make it easier for attackers to discover implementation errors.

[00:13:23] [SPEAKER_01] Now, that's an interesting argument. Because in technology, we often associate sophistication with improvement. More components, more automation, more elaborate protocols, more layers of control. But complexity is not the same thing as security. And simplicity is not automatically the same thing as safety.

[00:13:53] [SPEAKER_01] We need to look at both sides of the equation on both sides. And we got to start with the basic problem. The basic problem is that Bitcoin custody is fundamentally about controlling the cryptographic keys that authorize transactions. If someone obtains the ability to sign a valid transaction, that may be able to move the assets then. And depending on the circumstances, there may be no practical way to reverse that transaction.

[00:14:24] [SPEAKER_01] That makes custody a very different responsibility from managing an ordinary software application. In traditional financial systems, unauthorized transactions may be subject to intervention, dispute procedures, or recovery mechanisms. But those mechanisms vary and they are not guaranteed to exist. But Bitcoin's transaction model places enormous importance on preventing unauthorized signing in the first place.

[00:14:52] [SPEAKER_01] That's why institutions invest in specialized hardware, distributed authorization systems, code storage, and operational control. Multiparty comp, computation, and PC can allow several participants to contribute to a cryptographic operation without reconstructing a complete private key in one place. That can reduce certain risks.

[00:15:22] [SPEAKER_01] That can reduce certain risks, but implementing the system's security is not a trivial matter. The protocols, the software, they must be correct. The devices must be protected. The people administrating the systems must follow appropriate procedures. And the surrounding infrastructure must not resist compromise. Back's concern is that when too many moving parts are involved, the number of potential failure points increases.

[00:15:54] [SPEAKER_01] And that's a legitimate engineering concern. But here's what I want to challenge. That an offline signing system may reduce exposure to remote attacks, but that does not eliminate the need for operational controls.

[00:16:11] [SPEAKER_01] Somebody still needs to authorize transactions to manage access, to handle recovery, to ensure that the hardware is maintained, replaced when it's needed to be replaced, and protected against physical compromise. And someone needs to determine what happens when people responsible for signing disagree become unavailable or make mistakes.

[00:16:41] [SPEAKER_01] Security is not just a cryptographic problem. It's an organizational problem. You can have exceptionally strong cryptocurrency or cryptography supporting exceptionally weak governance. Or you can have sophisticated hardware protecting a process that authorizes the wrong transaction or the wrong people. You can have multiple signatures and still have inadequate separation of duties.

[00:17:10] [SPEAKER_01] So when Blockstream describes this new architecture, I think the appropriate response is interest combined with scrutiny. What specific threats does it mitigate? What assumptions does it make? How does it perform under stress? What independent testing has been completed? What happens during disaster recovery?

[00:17:36] [SPEAKER_01] How do institutions evaluate the security benefits against the operational costs? Those are questions that matter more than the elegance of the product announcement. There's another element, too, as well. Back pointing to the reported multi-billion dollar finance security incident. He did.

[00:18:01] [SPEAKER_01] As one example of the danger associated with compromised devices and access credentials. He's right. That illustrates the broader importance of protecting transaction authorization infrastructure. We should not assume that a different custody architecture would necessarily have prevented a particular breach without understanding the complete attack path. Security failures are often combinations of technical weaknesses.

[00:18:30] [SPEAKER_01] People, our decisions, and our organizational processes, which in a lot of startups and entrepreneurs and growing companies are overlooked or you don't have the people with the experience enough to understand the nuances of organizational processes and why they're important. It's a problem. It's a problem. It's a challenge. It's a problem.

[00:18:58] [SPEAKER_01] Changing one component does not automatically eliminate every other vulnerability and the AI itself argument also deserves attention because AI might help attackers identify software defects, analyze protocols, and automate aspects of exploitation.

[00:19:19] [SPEAKER_01] But AI can also assist defenders with testing, code review, monitoring, and incident response. The question is not simply whether attackers will have better tools. It's whether security organizations can adapt their methods quickly enough to manage the changing threat environment to keep ahead of the attackers.

[00:19:46] [SPEAKER_01] I think that's the broader lesson here, right? I think that's the broader lesson here, right? Because a Bitcoin custody is often described as ownership. But ownership without the ability to protect, recover, and responsibly administrate our assets can become a liability. The strongest security model is not necessarily the one with the most sophistication in the technology. And it's not necessarily the one with the fewest components.

[00:20:16] [SPEAKER_01] It's the one whose risks are understood, whose controls are effective, and whose operation remains dependable under realistic, unrealistic, any conditions. The lockstream is presenting one approach. That approach deserves examination. That approach deserves examination. But the real measure will be whether that approach demonstrates durable security in practice.

[00:20:46] [SPEAKER_01] Not simply whether its architecture sounds convincing.

[00:20:57] [SPEAKER_00] Segment three, when OpenAI enters payments.

[00:21:03] [SPEAKER_01] When AI moves from advice to action. On Friday, November 13, OpenAI introduced a payment wallet that allows eligible users to authorize purchases through chat GPT using existing payment networks. That wouldn't have been a good thing on Thursday. Maybe. I don't know. Initiative involves Visa, Mastercard, and MX, who are all decentralized protocols?

[00:21:33] [SPEAKER_01] I don't think so. Let's see what happened here. According to the announcement, the system uses tokenized payment credentials and biometric authorizations, such as facial recognition or fingerprint verification, to approve transactions. Maybe it doesn't use retina scans, but I remember some of the biggest opponents of, you know, was it WorldCoin? What is it? What was the OpenAI iris? Whatever it was.

[00:22:04] [SPEAKER_01] Being these banks, being, you know, payment providers yelling about it. Interesting. The initial rollout is limited to eligible chat GPT Pro customers in the US. I think I have Pro. I have Pro. And now, individual purchases require user approval. I think I have Plus, actually. No problem. I can't do it.

[00:22:28] [SPEAKER_01] OpenAI has also described plans to support future transactions within predefined spending limits and other restrictions, potentially allowing certain purchases to proceed without separate approval each and every time. That's good. That's good. I think. That future capability is important to distinguish from what has actually launched. The story is not simply about adding a checkout feature to a chat bot.

[00:22:57] [SPEAKER_01] It's about the relationship between human intent, human people, and machine execution. Until recently, much of the consumer AI experience involved asking questions and receiving responses. You might ask for product comparisons or travel recommendations or research a service or, but completing a transaction generally required moving into another system and making the purchase yourself.

[00:23:25] [SPEAKER_01] Now, those boundaries are changing. The same interface that helps you decide what to buy may also participate in executing the purchase itself. That can be rather convenient, but it can also reduce friction and can simplify purchasing processes, but it also opens up a new set of questions and possibly more friction because let's look at authorization.

[00:23:56] [SPEAKER_01] If I ask an AI assistant to find a suitable product, I've not necessarily authorized it to buy that product. If I tell it I have a budget of $500, that doesn't necessarily mean it can spend the $500 without consulting me. And if I say that I prefer one merchant over another, that doesn't automatically authorize substitutions when the preferred product is unavailable.

[00:24:23] [SPEAKER_01] Now, these distinctions may sound obvious, but they become extremely important when software moves from generating recommendations to initiating financial actions. Especially when it's noted that on right on the site, chat GPT can make mistakes. Claude can make mistakes.

[00:24:52] [SPEAKER_01] AI can make mistakes. If an AI system needs to understand the difference and does between a preference and instruction and a binding authorization. And the user needs to understand what the system has permission to do.

[00:25:11] [SPEAKER_01] The current requirement for transaction by transaction approval helps preserve that distinction, but as these systems evolve, predefined permission will become more significant. Now, imagine authorizing an AI agent to buy ordinary household supplies within a monthly budget. Okay, that's perfectly reasonable, right? Now imagine the agent selecting the wrong item. And that being the majority of your budget, then what?

[00:25:41] [SPEAKER_01] Or buying from a merchant whose return policy you never would have accepted, you can't charge back. Or making several purchases that individually comply with your spending limit, but collectively violate your intentions. What if they buy Scott instead of Charmin is asking?

[00:26:10] [SPEAKER_01] Or what if the AI misunderstands a conversation and treats the hypothetical example as an instruction? These are not necessarily reasons to prohibit automated purchasing. They're reasons to design authorization very carefully. We also need to examine accountability because suppose an AI agent completes a transaction that the user believes was unauthorized. Who is responsible? The AI provider? The payment network?

[00:26:41] [SPEAKER_01] The merchant? The bank that issued the card? Or the user who granted the agent certain permission? Huh. Existing payment systems already have rules for disputes, fraud, and unauthorized transactions, but AI-assisted purchasing may introduce ambiguity about what the customer actually intended to buy.

[00:27:07] [SPEAKER_01] And that's a problem that technology itself may not be able to resolve. The use of tokenized card credentials is a meaningful security measure because it can reduce the exposure of underlying payment information, but tokenization addresses only part of the risk. A transaction can be technically secure and still be undesired. The correct card can be charged for the wrong purchase.

[00:27:36] [SPEAKER_01] The authorized agent can misunderstand the authorized task. Security and consent are related, but they are not interchangeable. There's also a competitive dimension here too. If an AI system becomes the interface through which consumers discover and purchase products, it may influence which merchants are considered, which products are recommended, and how alternatives are presented,

[00:28:03] [SPEAKER_01] which raises questions about transparency. Are recommendations based primarily on user-stated preferences? Are commercial relationships involved? Can the user understand why one option was selected over another? And how easily can the user compare alternatives outside the AI system? Those questions become more important as AI interfaces assume a greater role in commercial decisions.

[00:28:35] [SPEAKER_01] I also think it's worth examining what this means for crypto itself. The initial open AI wallet relies on established card networks. It does not require consumers to use a blockchain or a stablecoin. Meanwhile, companies in the crypto industry are developing their own infrastructure for merchant-initiated payments, including stablecoin-based systems.

[00:29:02] [SPEAKER_01] These approaches may eventually compete, or they serve different purposes. Traditional card networks already offer extensive merchant acceptance, consumer protections, and established financial relationships.

[00:29:19] [SPEAKER_01] Blockchain-based systems may offer advantages in programmability, settlement, and certain machine-to-machine transactions, but neither approach should be assumed superior in every circumstance. The appropriate infrastructure depends on the use case. What matters here is that AI is beginning to participate more directly in economic activity.

[00:29:48] [SPEAKER_01] And that creates a responsibility to distinguish between what a machine can execute and what a person has authorized. The most important question may not be whether AI can make a payment. Clearly, we are moving in that direction. The more important question, however, is whether people will retain meaningful control over our decisions that are made on our behalf by these systems.

[00:30:19] [SPEAKER_01] Because convenience is valuable, but convenience should not make consent ambiguous. And automation should not make accountability disappear. Segment four, prediction markets again.

[00:30:45] [SPEAKER_01] Prediction markets and government officials when information becomes a financial advantage. The fourth story comes from Washington. I think that's the difference between what I did on Friday and today's. The other one was from Chicago, I think. November 13, the House Oversight Committee released findings from an investigation into the risks of government officials trading on prediction markets.

[00:31:11] [SPEAKER_01] Committee chairman James Comer called for Congress to consider additional restrictions, including the possibility of prohibiting certain government personnel from trading on these platforms. I think it's a good idea. Let's see. Investigation began in May and examined approximately 1,700 pages of documents. The committee also received five bipartisan staff briefings from prediction market companies.

[00:31:36] [SPEAKER_01] The concern is that government officials, military personnel, congressional staff, or individuals holding security clearances may have access to non-public information that could give them an unfair financial advantage. We don't like that. The committee suggested that Congress consider restrictions on certain contracts or potentially broader prohibitions. Now, let's look at these distinctions carefully here.

[00:32:06] [SPEAKER_01] First, prediction markets are not inherently insider trading operations. They allow participants to trade contracts tied to the outcomes of future events. Those events can involve economics, elections, government decisions, sports, and many other subjects. The prices of these contracts can reflect the market's collective assessment of the likelihood of an outcome.

[00:32:31] [SPEAKER_01] The fact that a market produces information does not mean every participant acquired information fairly. That's an issue. Consider someone who works inside a government agency. That person may learn about a decision long before it becomes public. Perhaps information concerns are an upcoming policy announcement, a regulatory action, or another event with a clearly defined outcome.

[00:33:01] [SPEAKER_01] If the person trades on a prediction market using knowledge, that knowledge, that information, the concern is not simply that they made a correct forecast. The concern is they may have used privileged access obtained through public service for private financial benefit.

[00:33:31] [SPEAKER_01] The concern is they may have used, they may have used privileged access obtained through public service. It creates, you know, that conflict, right? Between public responsibility and personal interest.

[00:33:51] [SPEAKER_01] First, we already recognize similar concepts in TRAFI, in the markets, inside of trading restrictions, but also ethics rules, disclosure obligations, conflict of interest policies. These all exist because access to material non-public information can create opportunities for abuse. Abuse.

[00:34:17] [SPEAKER_01] Prediction markets introduce additional questions because the underlying contracts may concern government actions rather than conventional corporate securities. To me, the person who penned the phrase blockchain ethics back in 2018, the ethical concern is recognizable. Public authority should not become a private trading advantage.

[00:34:46] [SPEAKER_01] That doesn't automatically settle what the law should require because there are several possible approaches here, legally. Congress could prohibit government personnel from trading on certain categories of contracts. It could establish broader restrictions for people with access to sensitive information. It could require disclosures, additional disclosures. It could strengthen monitoring and enforcement.

[00:35:13] [SPEAKER_01] Or it could consider comprehensive prohibitions for particular groups of officials. It might be in favor of that. Each approach has trade-offs. An narrow restriction may be easier to justify but harder to enforce. A broad restriction may be simpler to administer but could cover activity unrelated to an official's responsibilities.

[00:35:40] [SPEAKER_01] Disclosure may improve transparency without necessarily preventing misconduct. And enforcement depends on whether regulators and platforms can obtain sufficient information to identify violations. Government employees not automatically engage in misconduct simply because they participate in a prediction market.

[00:36:05] [SPEAKER_01] And an unusually successful trade is not by itself proof that someone used confidential information. I'm sure there are some Congress people who are just naturally gifted traders. I'm certain of it, right? Let me know your thoughts. Investigations need evidence. Allegations need to be tested. And restrictions need clear definitions.

[00:36:33] [SPEAKER_01] Otherwise, a legitimate effort to protect public trust can produce uncertainty and inconsistent enforcement. The committee's findings also raised questions about the responsibilities of prediction market platforms. The two biggest ones that we know of, Calci and Polymarket, have described steps intended to reduce insider trading risks.

[00:36:58] [SPEAKER_01] And that is relevant, but platform controls need to be evaluated according to what they can detect and prevent. My questions are here, or how do platforms identify restricted participants? How do they detect suspicious trading patterns? What information can they share with authorities? What happens when trading occurs through accounts that obscure the beneficial owner?

[00:37:27] [SPEAKER_01] And how do these controls work across different jurisdictions? Those are difficult questions. They're also familiar questions in financial regulation. What makes this issue particularly important is the relationship between government information and public trust.

[00:37:45] [SPEAKER_01] Citizens, you and I, should be able to expect that public officials are making decisions according to their responsibilities, not according to the potential value of their personal market positions. And that expectation should apply consistently. Not just to Republicans or Democrats or independents, not just to career civil servants or political employees or military personnel.

[00:38:15] [SPEAKER_01] The relevant standards should not change depending on which political party happens to be in control. I also want to distinguish the committee's recommendations from enacted law. A congressional investigation is not a conviction and a proposed prohibition is not an existing prohibition. And the release of findings does not establish that every concern raised has been legally proven.

[00:38:40] [SPEAKER_01] The next steps will involve determining whether Congress pursues legislation, what restrictions might be adopted, and how any resulting rules could be enforced. For now, the significance of this story is that prediction markets are becoming important enough to raise questions traditionally associated with financial market integrity and government ethics. That's not necessarily evidence that markets have failed.

[00:39:09] [SPEAKER_01] It's evidence that their growth is creating responsibilities that cannot be ignored. And the central question is really straightforward. How do we preserve the informational value of prediction markets without allowing the privileged access to public information to become a source of private profit?

[00:39:31] [SPEAKER_00] That's a question worth answering carefully now, isn't it? Segment five.

[00:39:47] [SPEAKER_01] Three central banks test the future of cross-border settlement. So we leave the U.S. and we go to international finance. November 13, the Bank of England, England, the Monetary Authority of Singapore, and the Bank of Thailand.

[00:40:09] [SPEAKER_01] And they announced the results of Project Meridian FX APAC, Asia Pacific. The project tested how traditional payment infrastructure could interact with distributed ledger technology to coordinate foreign exchange settlement. The Reserve Bank of Australia and the Central Bank of Philippines participated as observers.

[00:40:38] [SPEAKER_01] The experiment used a prototype developed by Partior to coordinate settlement across different systems, currencies, and time zones. One of the central concepts was payment versus payment settlement. Now that may sound technical, but the underlying problem is straightforward. When two parties exchange currencies, each party needs confidence that it will receive the currency it is owed. If one side pays and the other fails to deliver, the first party can suffer a loss.

[00:41:09] [SPEAKER_01] Payment versus payment mechanisms are designed to reduce the risk that by coordinating the exchange so that both sides settle together. Project Meridian explored how this coordination could work across traditional real-time gross settlement systems and distributed ledger environments. The prototype type matched settlement. The prototype type matched settlement instructions, coordinated the reservation of funds, monitored readiness, and initiated settlements.

[00:41:40] [SPEAKER_01] The project also examined multilateral netting. The multilateral netting is a process through which multiple payment obligations can be offset against one another, potentially reducing the amount of liquidity required to complete the settlement.

[00:41:58] [SPEAKER_01] The immediate significance is that the experiment demonstrated technical possibilities, but the more interesting story concerns the relationship between the established financial infrastructure and emerging technologies. For years, discussions about blockchain and tradfi have been framed as a competition.

[00:42:23] [SPEAKER_01] One system replacing the other, one model becoming obsolete, one technology eliminating the need for existing institutions. But financial infrastructure doesn't always evolve through replacement, and sometimes it evolves through integration. And that's what this project was exploring. Traditional real-time gross settlement systems are deeply embedded in national payment infrastructures.

[00:42:52] [SPEAKER_01] They support the settlement of significant financial obligations. They operate within an established legal, regulatory, and institutional frameworks. Distributed ledger systems offer different technical approaches to recording, coordinating, and transferring value, but connecting those systems is not as simple as establishing a software interface. There are questions of financial, of legal, legal finality.

[00:43:21] [SPEAKER_01] And questions of governance, operational resilience, and liability, and about how information can be shared across national borders. The project report acknowledged several of the challenges. And that acknowledgment matters because demonstrating that a prototype coordinate transactions does not mean that the resulting system is ready for widespread production use.

[00:43:49] [SPEAKER_01] A controlled experiment is not the same thing as a live financial network. The conditions are different. The participants are different, and the consequences of future and failure are different. And the obligations of the institutions involved are also different. And so the settlement finality as an example.

[00:44:11] [SPEAKER_01] When a financial transaction is considered final, the parties need legal certainty about whether the settlement can be reversed, challenged, or interrupted. A certainty is not determined solely by whether a blockchain records a transaction depends on the legal framework governing the assets, participating institutions, and the settlement arrangements. A technically completed transfer and a legally final settlement are not necessarily the same thing.

[00:44:42] [SPEAKER_01] Now consider governance, right? If several national payment systems interact through a common coordination mechanism, who is responsible for operating that mechanism? Who sets the rules? Who resolves the disputes? Determines what happens when one system becomes unavailable? And who bears a responsibility when a transaction fails after one participant has already taken an action?

[00:45:09] [SPEAKER_01] Those questions become particularly complicated when the system operates under different national laws. And that's why cross-border financial infrastructure is difficult to modernize. The technical problems do matter, but the institutional problems matter just as much. It's also a liquidity dimension. You know, banks and financial institutions must maintain sufficient funds to settle obligations.

[00:45:37] [SPEAKER_01] If better coordination and netting reduce the amount of liquidity, that could improve efficiency, but the benefits depend on how the system operates in practice. It depends on the transaction volumes, participation, timing, and behavior of institutions, and whether the new arrangements introduce costs or risks that offset the expected improvements.

[00:45:59] [SPEAKER_01] That's why I would resist calling this project a transformation of global finance to experiment with promising implications. Not proof that the global system has been reinvented. The Bank of England has indicated that it plans to introduce a live synchronization interface in 2028, which gives us a future milestone to watch for.

[00:46:24] [SPEAKER_01] But the existence of a roadmap does not guarantee that every capability tested will be implemented in production. There's still work to be done. And that brings us to the larger questions. What does financial innovation actually require? Does it require every institution to abandon its existing system? Does it require every asset to become a token?

[00:46:51] [SPEAKER_01] Does it require every transaction to move onto a public blockchain? I don't think the evidence supports those kinds of universal conclusions. Different systems have different requirements. Different institutions operate under different responsibilities as well. And different technologies may be appropriate for different purposes.

[00:47:11] [SPEAKER_01] What matters is whether the resulting infrastructure improves settlement reliability, reduces risk, increases efficiency, and maintains clear accountability. The experiment is interesting precisely because it does not assume that innovation requires the complete elimination of existing infrastructure. Instead, it examines whether systems built under different technological models can coordinate effectively.

[00:47:40] [SPEAKER_01] And that may prove to be an important distinction for financial modernization. Not a replacement for its own sake and not tokenization for its own sake, but measurable improvements of how financial obligations are settled. The central banks have completed a technical experiment. The next challenge is demonstrating whether the model can deliver durable benefits under real-world conditions.

[00:48:08] [SPEAKER_00] That is where this story becomes worth following.

[00:48:18] [SPEAKER_01] And that brings us to the end of this Monday's Weekend Roundup, five stories, each raising its own questions. Microsoft outage reminds us that operational failures can have consequences far beyond the system where they begin. Adam Back's argument for simpler custody, security challenges the assumption that technological sophistication automatically produces better protection.

[00:48:43] [SPEAKER_01] OpenAI's payment wallet introduces new possibilities for AI-assisted commerce, along with important questions about authorization and accountability. The House oversight investigation examines the boundary between public reasonability and private financial knowledge and advantages.

[00:49:06] [SPEAKER_01] And Project Meridian offers an example of how established financial institutions are experimenting with new ways to coordinate settlement. None of these stories has a simple universal answer, and none should be reduced to a headline that tells us what to think.

[00:49:27] [SPEAKER_01] The purpose of examining them is to understand what happened, what remains uncertain, and what deserves our attention as these developments continue. And that's work we do here. Thank you for joining me for another edition of the X Factor Monday Weekend Roundup. I'm Jamil Hasan. I'll see you next time.

[00:49:48] .

Digital transformation broadcast network

Follow Us on LinkedIn

Follow us on LinkedIn and be part of the conversation!

Powered by