Quantum Hits the Credit Desk, Tokenised Stocks Pile Up on Base, and Regulators Finally Open the AI Hood
As of this week, the technology frontier in finance stopped being a slide deck. American Express ran a credit-default model on 101 IBM qubits, JPMorganChase did error-corrected portfolio optimisation on 12 logical qubits, and the US banking regulators lined up to ask who exactly is in charge of your bank's AI agents.
When quantum walked into the credit department
For most of the last decade, 'quantum in finance' sat in the awkward middle ground between physics PhD talks and vendor marketing. This week it moved onto trading floors.
At the IEEE Quantum Week conference on 20 October, teams from American Express, JPMorganChase and Intesa Sanpaolo published results that, taken together, reframe how banks should think about quantum from 'research grant' to 'procurement project.'
American Express reported that adding features derived from up to 101 IBM qubits to a credit default model improved performance by 9 to 41 basis points, albeit within the authors' own error bars. JPMorganChase ran a 12-asset portfolio problem on 12 error-corrected qubits (made from 97 physical ones) and scored 0.131 against 0.136 for the unprotected baseline. Intesa Sanpaolo's quantum-inspired scheduler cut the cost of cash in roughly 900 ATMs by 12.7 per cent over a year.
No single result is a knockout. None of the hardware runs showed an unambiguous win over classical methods. But collectively, they confirm something quietly important: banks are no longer only reading quantum research, they are producing it with their own money, their own data and their own balance-sheet problems.
Gartner says finance is about to outspend governments on quantum
The commercial implication arrived with timing straight out of a comms team dream. On 8 October, Gartner forecast that global quantum computing revenue will top $1 billion in 2027 and that the banking, finance and insurance sector will become the largest source of quantum spending from 2028 to 2030, overtaking the public sector.
The projected BFI figure for 2028 is $289 million, nudging ahead of the public sector's $280 million. That is a line in the sand. For nearly a decade, quantum has been a national-security story first and a commercial one second. From 2028, per Gartner, the dollar signs flip to banks.
The deal flow catches up
The capital markets agreed. IonQ priced a $2 billion equity offering at a premium to its prior close. Xanadu announced a $3.6 billion merger with Crane Harbor Acquisition Corp to list on Nasdaq and the TSX. Pasqal and Crédit Agricole CIB said they will integrate quantum processing units into banking operations by 2028. Kipu Quantum and NTT DATA are developing commercial algorithms for supply-chain and financial-risk modelling.
Any single announcement would be interesting on its own. In aggregate, they describe an industry rapidly building the sales motion, the hardware roadmap and the client reference list that an enterprise technology category requires.
Tokenised stocks pile up: Base, Securitize and the SEC letterbox
While quantum captured the headlines, the quieter revolution kept running on-chain.
Base, Coinbase's layer-2 network, added 26 tokenised stocks for 24/7 trading over the past week. OKX and Intercontinental Exchange jointly filed with the US Securities and Exchange Commission to launch tokenised US stock trading. Securitize launched tokenised stocks on Solana. Separately, ARK Invest moved its $1.3 billion venture fund onto Ethereum via Securitize, per a P2P.org DeFi roundup.
The numbers are becoming meaningful. CryptBull estimates stock-linked tokens deployed in DeFi hit roughly $262.4 million at the end of September, nearly 18 times the start of 2026. A TokenPost report on 18 October put the stablecoin market cap at $302.34 billion, DeFi total value locked at $88.49 billion, and tokenised real-world assets at $34.61 billion on-chain.
For traditional capital markets, the awkward question is no longer 'will tokenisation happen?' It is 'which parts of my book now have a 24/7 shadow market, and does my risk desk know about it?'
Standard Chartered sees $2 trillion in tokenised RWAs by 2028
Standard Chartered said this week it expects $2 trillion of tokenised real-world assets by 2028, matching stablecoin size at the same horizon, per Cointelegraph coverage. The bank also warned that regulatory uncertainty remains the biggest threat to the RWA sector, a point the US Blockchain Association echoed in a GENIUS Act comment letter to Treasury, asking for clearer rules on DeFi and stablecoin platforms before the 19 October consultation deadline.
Regulators open the AI hood
Even quantum could not quite steal attention from the week's loudest regulatory move. Per Reuters on 13 October, the Federal Reserve, the Office of the Comptroller of the Currency and the Federal Deposit Insurance Corporation stepped up scrutiny of banks' use of advanced AI, particularly generative models and more autonomous agentic systems. Supervisory focus, BusinessWorld added, lands squarely on cybersecurity, autonomous decision-making and whether humans can still shut the systems down.
Deutsche Bank did not help calm the mood, reporting via Crypto Briefing that AI inference costs are falling 13-fold a year, faster than any previous transformative technology. Cheap intelligence plus autonomous agents plus unclear accountability lines is exactly the mix regulators tend to flag.
World ID lands at Money20/20
The practical response was on display at Money20/20. On 19 October, World launched World ID for Banking and Finance, pitched squarely at deepfake fraud and AI-driven impersonation across customer accounts, employee workflows and AI agents, per PR Newswire. Expect many more such launches. The pitch 'prove the human on the other end of this AI pipeline' is going to be a growth category for the next two years at least.
What ties the week together
Three threads knit the week into a single story. First, quantum computing in finance has moved from research sponsorship to procurement budget. The question for a CIO at a tier-one bank is now less 'when will we do a pilot?' and more 'which vendor, which problem, and who owns the roadmap?' Second, tokenisation is scaling horizontally. Stocks, real-estate funds and venture-capital vehicles are all moving on-chain at the same time, often with regulated infrastructure vendors like Securitize and Archax carrying the weight. Third, the regulatory and identity layer is finally catching up. The Fed, OCC and FDIC are asking who is accountable for the AI agents making credit, trading and operational decisions, while identity providers like World rush to supply the plumbing banks will need to answer.
The frontier is no longer a far horizon. It is sitting on the credit desk, inside the fund administration stack and in the regulator's inbox. And it arrived faster than most 2026 budget cycles planned for.
