Emerging Technologies in Finance: What Is in Production and What Is Still a Pitch Deck

Most lists of emerging technologies in finance rank them by how exciting they sound: AI first, quantum second, blockchain somewhere in between. That ordering hides the only question that matters to someone with money in the markets: is the technology running in production, at scale, today — or is it a pilot with a press release? The changes that already touch every trade are the dull ones: settlement moved to one day, and payments that clear in seconds. The ones that dominate conference agendas are still measured in pilots.

This article classifies each technology by where it actually stands in September 2026, using one dated number from the operator, the regulator or a named survey. If the only number available is a market-size forecast, the technology does not get a status. It gets left out.

The scorecard: what is live, what is small, what is still a pitch

Three labels, applied strictly. In production at scale means it carries a large share of real activity in its market. Live but small means real money moves through it, but it is a rounding error next to the system it wants to replace. Pilot / pitch means trials, test trades or a rule that has not taken effect.

Technology	Status, Sept 2026	Latest dated number	Source
T+1 settlement (US)	In production at scale	Mandatory since 28 May 2024; nearly 95% of trades affirmed by 9 pm ET on trade date	SEC; SIFMA/ICI/DTCC, Sept 2024
Algorithmic execution	In production at scale	“Virtually ubiquitous” in US cash equities; ~62% of electronic interdealer Treasury volume from principal trading firms	SEC staff report, Aug 2020
Instant payments (FedNow)	Live but small	1,909 institutions listed (21 Sept 2026); 5.0 million payments in Q2 2026	Federal Reserve Financial Services
Stablecoins	Live but small (large inside crypto)	About $312 billion in circulation (24 Sept 2026)	DefiLlama; rwa.xyz
Tokenised Treasury and money market funds	Live but small	$14.96 billion across 25 products (25 Sept 2026)	rwa.xyz
AI and generative AI in financial firms	Mostly pilot; some at scale	Of 149 traditional institutions, 44% piloting, 24% scaling, 6% transforming (surveyed Oct 2025–Jan 2026)	Cambridge CCAF, April 2026
Tokenised stocks and Treasuries at DTC	Pilot	First production test trades July 2026; three-year pilot under SEC no-action relief	DTCC; SEC staff, Dec 2025
Quantum computing in trading	Pilot	One published trial: up to ~34% relative gain in bond fill prediction (Sept 2025), not in production	HSBC; arXiv 2509.17715
US open banking rule (CFPB 1033)	Pitch (legally stalled)	Finalised Oct 2024; enforcement enjoined 29 Oct 2025; first compliance date passed unenforced	CFPB; E.D. Kentucky
T+1 settlement (EU, UK)	Scheduled	Go-live 11 October 2027	EU Regulation 2025/2075; FCA

What is missing from the table matters as much as what is in it. Virtual-reality trading floors, “quantum-inspired” everything, metaverse asset classes and IoT sensor feeds for retail traders appeared on most 2023–2025 lists, including an earlier version of this page. None of them has a dated production figure from an operator or regulator, so none of them earns a row.

In production at scale: the plumbing nobody markets

T+1 settlement

The largest technology change in US markets in the past decade was a settlement deadline. The SEC adopted the rule on 15 February 2023 and it took effect on 28 May 2024: most US stock, ETF and corporate bond trades now settle one business day after the trade instead of two. Brokers, custodians and asset managers had to automate trade matching so that it finishes on the evening of the trade date.

The industry’s own after-action report, published by SIFMA, ICI and DTCC on 12 September 2024, gives the before-and-after. Nearly 95% of transactions were meeting the affirmation cut-off of 9 pm ET on trade date, up from 73% in January 2024. Settlement fail rates did not rise: the CNS fail rate averaged 2.12% in July 2024, in line with the T+2 era. NSCC’s clearing fund, the collateral members post against unsettled trades, fell by about $3.0 billion, or 23%. One day less of open exposure means less margin tied up.

Europe follows. Regulation (EU) 2025/2075 moves EU trading-venue transactions to T+1 from 11 October 2027, and the UK government has published draft legislation for the same date. The same report, however, warns that same-day settlement “is not simply the next step”.

The practical consequence: cash from a sale is usable a day sooner, and the gap between selling one fund and buying another is shorter. It changed nothing about returns. It did remove a day of counterparty risk from every trade in the system, without a single headline about a “revolution”.

Algorithmic execution

Algorithms routing, slicing and pricing orders is the oldest item on any list of emerging technologies in trading, and it stopped emerging years ago. The SEC’s staff report on algorithmic trading, dated 5 August 2020, describes algorithms as “virtually ubiquitous” across activities in the US cash equity market. The report does not put a single percentage on it. The hard figure it does cite is from the Treasury market: principal trading firms, which trade almost entirely by algorithm, accounted for about 62% of volume in the electronic interdealer Treasury market (Brain et al., 2018).

Figures of 60–80% of US equity volume circulate widely. We could not trace any of them to a regulator’s measurement; they come from broker and vendor estimates with differing definitions. The accurate statement is weaker and more useful: almost every order you send is executed, routed or filled by an algorithm, whether or not you use one yourself.

Open-source libraries and broker APIs have made it cheap for individuals to build a system, not to find an edge. Our review of Python backtesting libraries found that the most-recommended one has not had a commit since April 2023. Our test of which algorithmic strategies are actually profitable shows how often an edge disappears after costs. The algorithmic trading guide covers the five components every system needs.

Live but small: instant payments and tokenised cash

FedNow and instant payments

The Federal Reserve’s instant payment service launched in July 2023. The Fed’s participant list dated 21 September 2026 has 1,909 entries. The Richmond Fed counted 1,725 in the first quarter of 2026, equal to 19.7% of US banks and credit unions. So after three years, roughly four in five institutions are still not on it.

Usage is growing fast from a small base. FedNow settled 4,997,811 payments in the second quarter of 2026, 2.35 times the volume a year earlier and 83% more than in the first quarter. The growth came from smaller payments: the count rose 83% while the value settled rose only 1.3%, so the average payment fell from $99,414 in Q1 to $54,957 in Q2. For scale, the Richmond Fed puts The Clearing House’s private RTP network at about 128 million payments a quarter. FedNow’s Q2 volume is about 4% of that.

The practical consequence: instant payments are real infrastructure that works, which is why they are in this table and not in the pitch column. About four in five banks and credit unions are not yet connected, so whether you can use it depends on where you bank.

Stablecoins

Stablecoins are the largest form of tokenised dollars by a wide margin: about 21 times the tokenised Treasury market described below. DefiLlama counted about $312.5 billion in circulation on 24 September 2026, up 6.2% in a year and 1.81 times the level two years earlier; rwa.xyz, with a narrower definition, shows $306 billion. Inside crypto markets they are the settlement currency, which is why they sit at the border of “live but small”. Outside it, they are equal to about 3.9% of the $7.94 trillion held in US money market funds (ICI, week ended 23 September 2026).

The regulatory status changed more than the size did. The United States now has a federal framework for payment stablecoins, covered in our explainer on the GENIUS Act and stablecoin regulation. Central banks are running a parallel experiment, compared in our piece on CBDCs in 2026.

Tokenised Treasury and money market funds

This is the category most pitch decks lead with, and the one where the gap between narrative and number is widest. On 25 September 2026, rwa.xyz tracked $14.96 billion in tokenised US Treasury products across 25 funds and 83,706 holder addresses. That is about 0.19% of US money market fund assets: for every dollar in a tokenised Treasury fund, about $531 sits in a conventional money fund.

BlackRock’s BUIDL fund, launched in March 2024 and usually cited as the flagship, held $2.24 billion. It is now third on rwa.xyz’s list behind Circle’s USYC ($2.43 billion) and Ondo’s USDY ($2.26 billion). It is also volatile in a way a normal money fund is not. BUIDL shrank 18.9% in the 30 days to 25 September, and the category as a whole fell 7.2%.

Tokenisation has been pitched to investors before; the 2017 wave of initial coin offerings promised the same frictionless access. The difference this time is that the underlying asset is a Treasury bill held by a regulated manager. That makes the product real. It does not make it large.

The practical consequence: a tokenised T-bill fund earns roughly the T-bill rate minus its fee, the same as a conventional money fund. What it adds is round-the-clock transferability on a blockchain, usually only between whitelisted holders. That matters to a trading firm posting collateral on a Sunday and very little to someone saving for retirement.

Pilot: tokenised stocks at the market’s central depository

The more consequential tokenisation project is not a crypto fund. It is the Depository Trust Company, the central securities depository behind US stock and bond settlement. On 11 December 2025, SEC staff issued a no-action letter allowing DTC to run a three-year pilot in which participants can hold their entitlements to certain securities as tokens. Eligible securities are limited to Russell 1000 stocks, US Treasuries and ETFs tracking major indices. Tokenised entitlements get no collateral or settlement value in DTC’s risk management.

On 15 July 2026, DTCC announced that more than 30 firms had used tokens converted from DTC-held securities in production transactions, including a Treasury repo trade, equity delivery-versus-payment trades, securities lending and margin workflows. A full launch is planned for October 2026. No volume figures were published.

That makes it a pilot by our definition, and an important one. If tokenised stocks ever reach ordinary US brokerage accounts at scale, the central depository is a likelier route than a standalone crypto platform.

AI in financial firms: wide adoption, shallow deployment

“AI revolution” headlines usually rest on the share of firms “using AI”, which is high and not very informative. The Bank of England and FCA survey of 118 UK firms, published on 21 November 2024, found that 75% already used AI. Foundation models (the category that includes large language models) made up 17% of use cases. Only 2% of use cases were fully autonomous; 55% involved some degree of automated decision-making.

The more recent and larger survey separates experimenting from running. The Cambridge Centre for Alternative Finance’s 2026 Global AI in Financial Services Report surveyed firms between October 2025 and January 2026 and published in April 2026. Among 149 traditional financial institutions, 75% reported some use of generative AI. But asked how far their AI deployment had actually gone, 44% were piloting, 24% scaling and 6% transforming their business with it. The other 26% were exploring or barely using it. Seven in ten were at the pilot stage or earlier. Fintechs were further along, with 19% at the transforming stage. Responses are self-reported from a voluntary sample, which may flatter adoption.

The practical consequence: when a product claims AI makes its investment decisions, the base rate says most financial firms are still testing where AI fits. Where firms report the biggest benefits so far, it is in data and analytics, anti-money-laundering and fraud prevention, and cybersecurity, not autonomous portfolio management. Our guide to AI trading agents covers what autonomy means in practice and where it fails. For investors who want exposure to the build-out itself rather than to AI-driven trading, AI ETFs are the listed route, and their holdings deserve a look before the label.

Quantum computing: one credible trial, no production

Quantum computing is the clearest case of a technology whose finance coverage runs years ahead of its evidence. The strongest public result so far came from HSBC and IBM on 25 September 2025. Using IBM Heron quantum processors on real, production-scale data from the European corporate bond market, HSBC’s team reported up to a ~34% relative improvement in predicting whether a quote would win the client’s trade, compared with standard classical methods.

The accompanying paper (arXiv 2509.17715) is careful in a way the headlines were not. The gain appeared only when quantum hardware transformed the data. The same transform run on a noiseless quantum simulator did not produce it. The authors write that “the inherent noise in current quantum hardware contributes to this effect and motivates further studies”. An improvement that depends on hardware noise, in one dataset, on a backtest, is an interesting research result. It is not a production trading system, and HSBC’s own release describes a trial, not a deployment.

Where quantum does affect finance today is defence. On 13 August 2024, NIST published the first three post-quantum cryptography standards (FIPS 203, 204 and 205). They are designed to protect encryption against future quantum computers. Moving bank and market systems onto them is a multi-year migration, and it is the quantum-related work finance actually has to do now.

Open banking in the US: a rule that exists on paper

Open banking, the right to move your financial data to another provider through a secure interface, is live in the UK and the EU. In the US it has been a regulatory pitch for two years. The Consumer Financial Protection Bureau finalised its Section 1033 rule in October 2024, with the largest institutions due to comply by 1 April 2026. A Kentucky bank and a bankers’ association sued the day it was issued. The Bureau, under new leadership, opened a reconsideration on 22 August 2025, asking among other things whether banks should be allowed to charge fees for data access. On 29 October 2025, the federal court in the Eastern District of Kentucky enjoined enforcement of the rule. The April 2026 deadline passed with nothing to enforce.

A replacement proposal was sent for White House regulatory review in August 2026. As of 25 September 2026, we could not find it published in the Federal Register. Data sharing between US banks and apps continues through private agreements; the federal right the 2024 rule promised does not yet exist in practice.

How to read the next “emerging technologies in finance” list

Three questions sort any claim quickly:

• Is there a dated number from the operator or regulator? A 2030 market-size forecast is not evidence of anything happening today.
• What is the number as a share of the system it replaces? $15 billion of tokenised Treasuries sounds large until it is set against $7.94 trillion in money funds.
• Does the source distinguish piloting from running? “75% of firms use AI” and “6% have transformed with it” can both describe the same industry.

By those tests, the technologies that have changed the cost and risk of an ordinary investor’s trades are T+1 settlement and algorithmic execution. Instant payments and tokenised cash are real and growing. Quantum trading and autonomous AI portfolio managers are still where the pitch decks are.

Frequently asked questions

What are the emerging technologies in finance?
The main ones in 2026 are faster settlement (T+1), instant payments, stablecoins, tokenised funds and securities, artificial intelligence, quantum computing and open banking. They are at very different stages: T+1 and algorithmic execution are in production at scale; instant payments, stablecoins and tokenised Treasury funds are live but small; quantum computing and US open banking are still pilots or stalled rules.

What emerging technologies are changing the financial services industry?
The measurable changes are in back-office plumbing: one-day settlement in the US since May 2024, instant payments through FedNow and RTP, and AI for data analytics, fraud and anti-money-laundering checks, and cybersecurity. AI is widely tried but less widely scaled: in a 2026 Cambridge survey, 30% of traditional financial institutions had reached the scaling or transforming stage.

What are the emerging innovations in financial technology?
The innovations with real money behind them are stablecoins (about $312 billion in circulation in September 2026), tokenised Treasury funds (about $15 billion) and instant payment rails. DTC’s tokenisation pilot for Russell 1000 stocks, Treasuries and index ETFs is the most significant development to watch, with a full launch planned for October 2026.

How is technology used in trading today?
Almost every order is routed and executed by algorithms. The SEC’s staff describes them as virtually ubiquitous in US cash equities. Trades now settle one business day after execution in the US. Fully autonomous AI decision-making remains rare: 2% of AI use cases in the 2024 UK regulators’ survey.

What is advanced trading technology for digital assets?
The most established pieces are stablecoins used as settlement cash and tokenised money market funds used as collateral, both available around the clock on blockchains. The newest is tokenised versions of conventional securities held at the central depository, which DTC began testing in production trades in July 2026 under a three-year SEC staff no-action pilot.

Which emerging technologies matter in asset management?
Tokenised fund share classes, AI for research and operations, and faster settlement. Tokenised Treasury funds from managers such as BlackRock, Franklin Templeton and WisdomTree hold about $15 billion combined, around 0.19% of US money market fund assets.

Is quantum computing used in financial trading yet?
Not in production. The most credible public result is HSBC and IBM’s September 2025 bond-trading trial, which reported up to a 34% relative improvement in predicting trade fills. The gain appeared only on noisy quantum hardware and not in a noiseless simulation, which the authors flag for further study. The quantum work banks are actually implementing is post-quantum encryption, based on NIST standards published in August 2024.

Sources

• U.S. Securities and Exchange Commission, “SEC Finalizes Rules to Reduce Risks in Clearance and Settlement”, press release 2023-29, 15 February 2023 (T+1, compliance date 28 May 2024). sec.gov
• SIFMA, ICI and DTCC, T+1 After Action Report, 12 September 2024 (affirmation rate, fail rates, NSCC clearing fund). sifma.org
• Regulation (EU) 2025/2075 amending Regulation (EU) No 909/2014 as regards a shorter settlement cycle, Official Journal, 14 October 2025, applicable from 11 October 2027. eur-lex.europa.eu
• Financial Conduct Authority, “About T+1 settlement”, updated 13 August 2026 (UK draft statutory instrument of 20 November 2025, go-live 11 October 2027). fca.org.uk
• SEC Division of Trading and Markets staff, Staff Report on Algorithmic Trading in U.S. Capital Markets, 5 August 2020, sections I.C, IV and V.B (citing Brain, Pinto, Choi and Nguyen, FEDS Notes, 2018). sec.gov (PDF)
• Federal Reserve Financial Services, FedNow Service Volume and Value Statistics, quarterly table, last updated 6 July 2026; and FedNow Service Participating Financial Institutions, list as of 21 September 2026. frbservices.org, participant list
• Wang, Z. and Phan, V., “FedNow and the Development of U.S. Fast Payments”, Federal Reserve Bank of Richmond Economic Brief No. 26-28, August 2026. richmondfed.org
• DefiLlama, stablecoin circulating supply, all pegs, daily series, 24 September 2026. defillama.com
• rwa.xyz, Tokenized U.S. Treasury Funds and Stablecoins dashboards, data as of 25 September 2026. app.rwa.xyz
• Investment Company Institute, Money Market Fund Assets, week ended 23 September 2026, released 24 September 2026. ici.org
• SEC Division of Trading and Markets, no-action letter to The Depository Trust Company regarding DTCC Tokenization Services, 11 December 2025. sec.gov (PDF)
• DTCC, “DTCC Turns Tokenization into Reality”, press release, 15 July 2026. dtcc.com
• Bank of England and Financial Conduct Authority, Artificial intelligence in UK financial services – 2024, 21 November 2024. bankofengland.co.uk
• Cambridge Centre for Alternative Finance, The 2026 Global AI in Financial Services Report: Adoption, impact and risks, April 2026, Figures 1.0 and 1.1. jbs.cam.ac.uk
• HSBC, “HSBC demonstrates world’s first-known quantum-enabled algorithmic trading with IBM”, 25 September 2025. hsbc.com
• Ciceri, A. et al., “Enhanced fill probability estimates in institutional algorithmic bond trading using statistical learning algorithms with quantum computers”, arXiv:2509.17715, 22 September 2025. arxiv.org
• NIST, “Post-Quantum Cryptography FIPS Approved” (FIPS 203, 204, 205), 13 August 2024. csrc.nist.gov
• Consumer Financial Protection Bureau, Required Rulemaking on Personal Financial Data Rights, final rule, October 2024; and Personal Financial Data Rights Reconsideration, advance notice of proposed rulemaking, Federal Register, 22 August 2025. consumerfinance.gov, federalregister.gov
• Consumer Finance Monitor (Ballard Spahr), “CFPB Sends New Section 1033 ‘Open Banking’ Proposal to OIRA for Review”, 6 August 2026 (secondary; used only for the OIRA submission). consumerfinancemonitor.com
• U.S. District Court, Eastern District of Kentucky, Forcht Bank, N.A. et al. v. CFPB, order granting preliminary injunction, 29 October 2025. court order (PDF)
• Ratios (shares of money market fund assets, growth multiples, FedNow versus RTP) are our calculations from the figures above.

Related reading

• AI trading agents in 2026: what autonomy means in practice and where it breaks.
• Which algorithmic trading strategies are actually profitable?: evidence, costs and a 33-year backtest.
• Algorithmic trading strategies for 2026: signal, sizing, execution, risk and monitoring.
• Python backtesting libraries in 2026: what the rankings get wrong.
• The GENIUS Act and stablecoin regulation: reserves, redemption and who may issue.
• CBDCs in 2026: the digital euro, the e-CNY and the stablecoin alternative.
• Artificial intelligence ETFs: what is actually inside them.

This article is general information and education, not personalised investment advice. It does not take into account the financial situation, objectives or risk tolerance of any individual reader, and it does not recommend any specific security, fund, token or service. Figures describe the dates stated and can change quickly, particularly for stablecoins and tokenised funds. Digital assets can lose value and may not carry the protections of conventional bank deposits or funds. Capital is at risk and past performance does not indicate future results.