When Fintech Meets AI: Trillion-dollar Opportunities for Global Finance—And Fraudsters

Artificial intelligence is structurally reshaping the model of modern fintech. Or rather, not AI itself, but the technologies its development spawns. As ever in a scientific and technological revolution, this brings both opportunities and systemic challenges.

Not least, this concerns the security of the financial system. Here AI opens up new avenues for defense—but also vast scope for those who want to “hack the system.”

It resembles the perpetual contest between the burglar and the engineers who design ever-new locks: for every new lock, the burglar readies a new pick.

But first, let us unpack how the new crypto-based international financial system works.

The old Bretton Woods model rested on a strong dollar and the dollar’s gold standard. The Jamaica currency system, which replaced Bretton Woods in the 1970s, was built instead around a pool of global reserve fiat currencies and a mechanism of market-determined exchange rates.

In reality, the global financial model still rested on that same dollar and on dollar-denominated instruments, above all Treasuries (US government bonds). The dollar drew added support from its status as the currency of oil settlements: the petrodollar.

In recent years, however, a number of the world’s central banks have been actively offloading Treasuries.

At the same time, they have been buying gold in earnest. In the make-up of their reserves, gold has overtaken Treasuries: according to the European Central Bank, by the end of last year gold had reached 27 percent of the foreign exchange reserves of the world’s central banks, while the share of Treasuries had shrunk to 22 percent. As for Treasuries themselves, China at its peak held a portfolio of these securities worth $1.4 trillion; today its reserves contain US bonds worth $640 billion. Several countries of the Global South—India, China, Brazil—are actively selling Treasuries.

The threat of major wars is pushing countries to accumulate gold and adjust their Treasury holdings.

Why central banks are dumping Treasuries and buying gold

The liquidity threat to the US financial system is plain enough: an outflow of global liquidity from dollar assets was observed at the height of Trump’s roll-out of import tariffs, for instance, and again at the peak of the war in the Middle East, when the Saudis and the UAE were selling down their portfolios of American bonds. How does the United States intend to offset the outflow of global liquidity from the dollar fiat system? Above all, by developing a mechanism of quasi-private money: stablecoins.

In other words, by building financial “links” between fiat and crypto. That is precisely why Trump sent Congress a raft of bills: on cryptocurrency issuance, on developing the mechanism of quasi-private money and on barring the Federal Reserve from issuing a digital dollar (lest it become a rival to stablecoins).

Howard Lutnick, the current commerce secretary in the Trump administration and owner of a financial firm that became legendary after 11 September 2001, has repeatedly stressed the need to build financial links between crypto and fiat and to develop stablecoins, which were to become the foundation of a new global financial system.

What a stablecoin is and how Tether works: USDT, EURT, CNHT, XAUT

A stablecoin is backed by an underlying asset. That may be the dollar, the euro, gold, the yuan or Treasuries (a dollar instrument).

Take the company Tether and its eponymous cryptocurrency stablecoin USDT, pegged one-to-one to the US dollar. It is one of the most striking financial links between fiat money and the cryptocurrency market.

Alongside its dollar-pegged token (USDT), the company issues stablecoins for other currencies—the euro (EURT) and the yuan (CNHT)—as well as a cryptocurrency backed by physical gold (XAUT). The value of Tether’s tokenized gold has now surpassed $23 billion.

One XAUT token equals one troy ounce of gold. Together with Ledn, the largest cryptocurrency lender, the company is launching a platform for issuing loans in fiat currencies against collateral in USDT and XAUT—that is, dollar and gold stablecoins.

Gold-backed lending used to be the preserve of central banks. Now the technology is going “to the people.”

The tokenization of national assets is becoming a key factor in global competitiveness.

Within five to ten years, a substantial share of private investment will flow into tokens backed by various kinds of assets in one country or another.

Foreign direct investment will give way to tokenized investment.

There is a lesson here for Ukrainian business too: if you want to attract investment, tokenize your assets.

But back to Tether. The company already holds 140 metric tonnes of gold, making it the largest corporate owner of the precious metal—ahead even of some countries, Ukraine among them.

Tether is tokenizing artificial intelligence and big data projects, as well as renewable energy ventures. The latter is an outright win-win model: investment in AI drives demand for clean energy.

Tether’s Treasury portfolio could overtake Japan’s and Britain’s by 2030

Companies like Tether are also becoming financial pumps channelling liquidity into the Treasury system—and this model does not depend on sanctions or on relations between states. To back its dollar stablecoin, the company has already built a Treasury portfolio of $117 billion, more than many countries hold in this asset.

By various estimates, the company’s Treasury portfolio could reach $1.4 trillion by 2030, overtaking the holdings of the current key investors in American debt—Japan and the UK. Nor does this pumping of liquidity into Treasuries depend on central banks’ interest rate policy.

When investors anticipate cuts to the Fed’s base rates, for instance, they all expect Treasury yields to fall and prices to rise, so they buy actively; and vice versa.

Tether, however, buys Treasuries in response to rising customer demand for its dollar stablecoin; the strategy of rising or falling yields affects its policy only indirectly.

In other words, no more carry trade, in which investors funded themselves in, say, Japan at close to zero and invested in American debt securities (for a long time the Japanese central bank’s rate was negative).

From now on, flows of liquidity into the American public debt system will not depend on other states’ monetary policy. A perfect model. Within it, a sell-off of Treasuries is possible only if holders redeem their dollar tokens.

In that case, customers would demand fiat dollars, and Tether would be forced to sell part of its Treasury portfolio to raise the necessary cash. But that could happen only in the event of a global digital and energy blackout striking the world’s big data—in which case everything would go down, not just Tether.

For now, the forecast runs as follows: by 2030, a substantial share of fiat currencies will migrate into stablecoins—that is, onto the blockchain—which, incidentally, will deal a serious blow to traditional systems of bank settlement.

Every large company will seek to tokenize its assets and issue its own stablecoin. CeFi and DeFi—centralised and decentralised financial systems—will merge.

This new profile of global fintech is essential for understanding the following trends:

1) the development of financial technology will be inseparably bound up with AI products and technologies—the age of fiat currencies is rapidly changing shape;

2) AI will be the “engine” of fintech’s new digital development model;

3) the use of AI will carry substantial risks for the security of transactions—challenges more complex even than call centers running phone scams;

4) the merger of CeFi and DeFi (centralized and decentralised financial systems) will remove the option of blocking external AI-enabled attacks within notionally isolated financial clusters;

5) AI could put attacks on fintech on an “industrial footing.”

What McKinsey says about AI in banking

In its analytical report on the application of AI in fintech, AI-powered decision making for the bank of the future, McKinsey notes:

“Banks are already strengthening customer relationships and lowering costs by using artificial intelligence to guide customer engagement.

Success requires that capability stacks include the right decision-making elements.

The ongoing transition to digital channels creates an opportunity for banks to serve more customers, expand market share and increase revenue at lower cost.

Crucially, banks that pursue this opportunity also can access the bigger, richer data sets required to fuel advanced-analytics (AA) and machinelearning (ML) decision engines.

Deployed at scale, these decision-making capabilities powered by artificial intelligence (AI) can give the bank a decisive competitive edge by generating significant incremental value for customers, partners, and the bank.

Banks that aim to compete in global and regional markets increasingly influenced by digital ecosystems will need a well-rounded AI-and-analytics capability stack comprising four main layers: reimagined engagement, AI-powered decision making, core technology and data infrastructure, and leading-edge operating model.”

Artificial intelligence can help detect fraudulent transactions, shape new algorithmic-trading technologies for financial instruments, and create a new standard of customer service.

McKinsey reckons AI could bring fintech roughly a trillion dollars of additional capitalisation every year.

Yet integrating AI into the core of fintech carries systemic cybersecurity risks.

A phenomenon of “notional fraud” may emerge here—for instance, the manipulation of trading algorithms in the interests of particular players. This would be, so to speak, “legal fraud.”

But classic fraud may intensify too: AI-based language models could lead to leaks of customers’ personal data, while cloud infrastructure could come under attack from AI-based hacking models. Complex, highly structured fintech cloud systems will be especially vulnerable.

AI’s opportunities in the fintech sector:

• machine-learning algorithms that analyze millions of transactions per second to detect anomalies and pre-emptively block suspicious operations;
• credit scoring: AI reduces banks’ repayment risk by weighing hundreds of parameters when assessing a client’s solvency;
• compliance (KYC): automated customer screening and verification of documents against databases.

Set against these, AI’s risks for fintech are:

• deepfakes and social engineering: fraudsters generate convincing voice messages and video calls—purporting to come from “the boss” or “a relative”, say—to trick people into transferring money;
• smart phishing: AI makes it possible to craft grammatically flawless, personalised phishing emails and malicious websites;
• attacks on algorithms: attempts to crack or deceive banks’ own security systems using other algorithms.

This contest between burglar and locksmith is, of course, an eternal one. No one is going to forgo the possibilities of AI, just as no one gave up the internet because of hacking attacks, or mobile phones because of call center scams.

At stake is a trillion dollars a year in additional capitalization—and nobody is going to walk away from that kind of money.