COOPER BECKWITH, LLCQUALITY · VALUE · SERVICE

Research

Papers on AI risk, agents, capital, and liquidity in financial institutions, written for the people who sign the numbers, not the people selling the tools. Published under the author's own byline from High Point, North Carolina.

Know Your Customer, Know Your Agent

Amex Bank Just Agreed to Pay $350MM for AML Risks It Didn't Know It Had. What Happens When the AI Agents Get Involved?

Know Your Customer, Know Your Agent, bonus paper · October 14, 2026 · John Beckwith

On October 8, 2026 the OCC found systemic breakdowns in American Express National Bank's suspicious activity monitoring. Its risk assessment watched deposit products while about $13 billion of suspected trade-based money laundering moved through card charges and repayments over eleven years. A BSA and AML program watches what its risk assessment tells it to watch, and AI agents are the next business most risk assessments have not seen.

Know Your AgentAML and KYCAI risk
Know Your Customer, Know Your Agent

Forty-Five Minutes, $460MM, and 97 Unread Emails

Know Your Customer, Know Your Agent, part 4: the developer and the deployer · October 12, 2026 · John Beckwith

On August 1, 2012, Knight Capital lost more than $460 million in about 45 minutes to trading code it had retired years earlier and never removed, while 97 automated warning emails reached staff who did not act on them. The firm had no procedure to stop its own machine. The agent you deploy is the agent you own, and contract law, a 2024 tribunal, and a bipartisan Senate bill announced October 1, 2026 all say so.

Know Your AgentAI governanceAI risk
Know Your Customer, Know Your Agent

The Credentials Were Valid. The Instructions Were Not.

Know Your Customer, Know Your Agent, part 3: the mandate · October 9, 2026 · John Beckwith

In February 2016, $951 million of fraudulent payment instructions sent from Bangladesh Bank's own SWIFT terminal passed authentication, and $81 million was stolen. Authentication proves who holds the key. Authorization proves the holder had authority for this instruction, at this amount, to this payee, today. A customer's AI agent will present perfectly authenticated instructions, and the control that tells a legitimate one from a stolen or mistaken one is a record of what the customer authorized: the mandate.

Know Your AgentPaymentsAI risk
Know Your Customer, Know Your Agent

The New Casas de Cambio

Know Your Customer, Know Your Agent, part 2: the operator · October 7, 2026 · John Beckwith

Between you and the person whose money moves there is now a platform that runs the agent: a frontier lab's assistant, a big tech wallet, a fintech, a data aggregator, or an agent sold through your core provider's marketplace. That platform is the operator. It identifies its own principals, aggregates thousands of them, and presents your institution with volume at machine speed. Wachovia paid $160 million in 2010 for treating $420 billion of casa de cambio traffic as if it came from customers it knew.

Know Your AgentAML and KYCAI risk
Know Your Customer, Know Your Agent

Who Is Behind the Agent at Your Window?

Know Your Customer, Know Your Agent, part 1: the principal · October 6, 2026 · John Beckwith

An AI agent can open a session at your institution with your customer's credentials, read the account, and move money, and to your systems it looks like the customer. In 2012 HSBC paid $1.92 billion for knowing its counterparty and not the principals behind the counterparty's traffic. The agent at your window opens the same gap.

Know Your AgentAML and KYCAI risk
Research paper

AI Risk in Depositories: The Need for Elicitability

Why community institutions should adopt AI deliberately · September 17, 2026 · John Beckwith

Artificial intelligence is real, compounding, and about to accelerate as models begin improving themselves. That is not a reason to run it into your books and records, and it is not a reason to stand in front of it. It is a reason to be deliberate. From 2006 to 2008, opaque models stacked on leverage produced numbers no one understood and losses no one could locate until it was too late, and community institutions carried real damage.

AI riskAI governanceBooks and records

Co-authored work

Papers written with colleagues and published by their firms. Summaries here, full text at the publisher.

Co-authored work

Why Rates Are High: Five Forces, Sized

October 13, 2026 · with Sanjay Sharma, published by GreenPoint Financial

An attribution of the 376-basis-point rise in the 10-year Treasury yield since the end of 2021 across sticky inflation, two wars, a $1.9 trillion deficit, and a Federal Reserve that buys only bills, starting from the market's own decomposition into expected inflation, the real rate path, and the term premium.

Co-authored work

How to Manage High Interest Rates: 10 To-Dos and 10 Not-To-Dos

October 13, 2026 · with Sanjay Sharma, published by GreenPoint Financial

A one-page-per-item operating list for institutions running long assets against short funding in the highest rate environment since 2007, written for boards and ALCO committees.

Co-authored work

Fixed-Rate Mortgages at 7%: 10 To-Dos and 10 Not-To-Dos

October 13, 2026 · with Sanjay Sharma, published by GreenPoint Financial

Half the country's mortgages are locked below 4 percent, community banks carry theirs at 5.9 percent, and new ones cost 7.4 percent. What extension risk does to a community bank's margin, with a coefficient, and what to do about it.

Co-authored work

Well-Capitalized on Wednesday. Closed on Friday.

September 30, 2026 · with Sanjay Sharma, published by GreenPoint Financial

Why sharp moves in rates and the yield curve are existential for community banks, credit unions, and nonbank lenders that do not manage both sides of the balance sheet. Marks community bank securities losses to late-September yields, traces the rate regime to technology capital spending and the federal deficit, and reads the 2023 failures as mechanism rather than history.

New papers by email

Free. Every new paper reaches subscribers first.