A library of snippets
xxx
Scams using AI generate an average of $3.2 million per operation, compared to $719,000 for those that don’t, according to crypto monitoring firm Chainalysis’ 2026 Crypto Crime Report. That nearly fivefold increase in revenue per scam is driven by AI’s ability to automate impersonation — a tactic that saw 1,400% growth this year, according to the company.
From: AI agents help uncover crypto crime | American Banker.
xxx
xxx
Founded by Circle co-founder Sean Neville, Catena is a banking and governance platform for AI agents, providing them with programmable financial infrastructure covering accounts, payments, stablecoin rails across ten chains, and yield on idle balances, accessed through an MCP server, a REST API, or a command line.
As it opens up private access to its platform, the firm has revealed that it has raised $30 million in a round co-led by a16z crypto and Acrew, with participation from Breyer Capital, General Catalyst, QED, Oak, Fin, and IDG Capital.
The money – which brings total financing to $48 million – will be used to build the platform, earn the licences, and hire the people required to serve agents and their operators.
Meanwhile, the Office of the Comptroller of the Currency has accepted for filing Catena’s application for a National Trust Bank charter.
From: Circle co-founder raises $30m for agentic finance startup Catena.
xxx
xxx
The GitHub repository, reportedly named “Private-CISA,” was public until this past weekend, according to KrebsOnSecurity.
It was tied to Nightwing, which reportedly was working with CISA and contained cloud keys, authentication tokens, plaintext passwords, internal documentation, deployment logs, and files that described how the agency builds, tests, and deploys software.From: GitHub leak exposed CISA, DHS GovCloud keys, internal credentials | Biometric Update.
xxx
xxx
Circle’s agent tools include Circle CLI, a command interface that lets developers and potentially AI agents build applications on top of Circle’s platform suite, with a focus on wallets, payments and policy management for agents. The other products include: Nanopayments, a protocol that supports USDC transfers designed for large volumes of machine-to-machine payments; Agent Wallets, which are designed for AI agents to hold, send, and manage funds autonomously within predefined guardrails; and Agent Marketplace, a directory of agentic services that both humans and AI agents can browse and integrate to enable agents to discover and pay for services programmatically.
From: How USDC issuer Circle is embracing agentic commerce | American Banker.
xxx
xxx
Cryptocurrency companies are preparing for the threat that powerful quantum computers could soon be able to hack the security at the heart of the global industry, including breaking the critical code that underpins bitcoin.
The risk to crypto posed by fast-developing quantum technology — which exploits the way the physics of matter works differently at atomic and subatomic levels — was once considered a distant possibility, with bitcoin widely seen as unhackable.
But digital assets firms are speeding up their preparations for a “post-quantum” age, as tech companies slash the timelines for developing practical quantum computers to as soon as 2030.
“The threat has moved from theoretical to credible,” said Ayo Akinyele, head of engineering at RippleX, the blockchain development arm of crypto group Ripple.
xxx
The brilliant author John Lancaster, in his review of Oliver Bullough’s fascinating new book on money laundering “They Want Our Dollars” makes a sweeping and damning indictment of our current efforts to combat this global menace, saying the sheer extent of the legal apparatus (which costs hundreds of billions of dollars around the world) juxtaposed with the sheer extent of the crime, represents failure on a colossal scale. Spot on.
(And this in a country where the late and unlamented Russian warlord Yevgeny Prigozhin passed his money laundering checks with his mother’s gas bill from St. Petersburg.)
xxx
If it were an industry, money laundering would be the third biggest business in the world, behind commercial property and ahead of pensions.
xxx
xxx
he AI Agent Mesh is the data plane where AI agents communicate with each other to complete tasks assigned to them
xxx
In the United Kingdom we are about to ask a hundred random people to design our national digital identity scheme. Seriously. Invitations went out to 36,000 UK addresses (mine wasn’t one of them) asking recipients to comment on “How should we design a Digital ID system for the UK?”. Of those who respond, the government will choose (via some opaque process) somewhere between 100 and 120 of them to come to central Birmingham for a couple of weekend sessions (plus some Zoom sessions in the week). The government says that those chosen will represent “a cross-section of the UK population”. Who knows what they will come up with.
Meanwhile, in advanced nations, life goes on. China has just announced that robots will be getting digital identities and launched the Humanoid Full Lifecycle Management Service Platform to manage them. The identities will be used to track the robots from assembly through to recycling in a programme led by the Humanoid Robotics and Embodied Intelligence Standardization (HEIS) committee, part of the country’s Ministry of Industry and Information Technology.
That fact is that the demands of the economy mean that digital identities not only for people but for businesses and bots are a priority, a foundational layer for not only the economy but for a safe and secure society. If we want economic growth, we have get this layer in place.
Noelle Acheson, who knows her onions* when it comes to real-world markets, says “It’s unlikely a tokenized share in Palantir will be able to ‘seamlessly’ plug into a Russian DeFi application, for instance, or any DeFi application that has no KYC unless we finally sort out the onchain identity issue”. I am a big believer in digital assets and tokens and next-generation financial market infrastructure (FMI) and all that jazz but Noelle is right.
In the Bank for International Settlement (BIS) white paper on the topic of next generation FMI, they themselves frame digital identity as a fundamental platform on which the financial internet (that is, the “finternet”) needs. Until there is a working identity infrastruture in place, the finternet cannot happen.
Now, that identity infrastructure does not need to a global ID solution used for everything or network of national IDs or anything like it. It might well be an identity infrastructure solely for financial services. Some kind of financial services passport, for want of a better term. I think that one way we can make some progress is by focusing on a practical solution here rather than try to boil the ocean and solve the entire world’s superset of (wholly different) identity problems.
What if financial instutitions got together to solve this problem for themselves?
It is a fairly straightforward idea, after all. The first time you come into contact with the financial system, you are KYCd, AMLd, CTFd and PEPd up the wazoo: folded, spindled and mutilated until you are a certified player, at which point you would be issued with a financial services passport that you can use to visit other institutions, accountants, lawyers and so on. Then, instead of sending your personally-identifiable information with a transaction, you need only send a pointer to the passport (a public key, in essence). There will be a visa to the land of (eg) Barclays stamped inside the passport (a verifiable credential, in essence) which testifies to your certified playerness. You can show this visa to anyone: they can’t copy it our counterfeit it because they do not have access the private key, which is locked away in secure hardware somewhere.
In one of their reports last year, McKinsey said that what they label “credentialling and identity” is the first of their key control points in the agentic economy becaue agents need secure, user-granted permission before they can initiate transactions across multiple institutions. Indeed they do and therefore, organisations that already manage high-trust credentials start with a clear advantage. They go on to highlight some success factors: zero-trust architectures that never assume persistent access, dynamic consent via standardized protocols (for example, OAuth2/OpenID Connect) and continuous audit trails.
All of these ought to be
* I deliberately used onions here, rather than apples or SEC rules, because the trading of onion futures in the US is illegal, a fact that to my mind at least highlights just why markets are not just about technology.
xxx
The eMarketer analysis of Stripe’s annual letter notes that in 2026, 95% of e-commerce sales driven by AI platforms will still be completed off the AI platform and on the merchant site. The agent drives discovery. The human still clicks “buy.”
From: (1) Simon Taylor on X: “The 5 Levels of Agentic Commerce” / X.
xxx
Have a WordPress.com account? Log in now.
