Erica Stanford, an adviser at CMS UK and author of Crypto Wars, told the Payment Expert Podcast that AI has lowered the barrier to payments fraud, that stablecoin payment volumes are far smaller than headline figures suggest, and that crypto is now more traceable than cash.
Stanford works in a non-legal capacity on the international law firm’s crypto team.
Stablecoins: hype versus payments
Stablecoins settled more on-chain value last year than Visa‘s own network processed. Stanford said the headline numbers overstate how much equates to real payments.
The roughly $13trn quoted for total stablecoin volume “includes a lot of automated and bot payments,” she said; stripping those out leaves “just over $3trn of payments a year,” and about $390bn in pure payments.
Much of the rest is crypto trading, where stablecoins are “the main settlement between different cryptocurrencies.”
As payments, “they’re intended to be digital money.” In Latin America, people rely on them “because some local currencies aren’t stable.” The US GENIUS Act bars issuers from paying yield, and Stanford said yield is not the primary appeal, because users care more whether they can “pay staff, contractors or gig workers, or receive remittances in a stable form of value.”
Crypto still carries a money-laundering reputation among banks and regulators. Stanford said the data does not support it, though. In Bitcoin‘s first couple of years “about 30% of all Bitcoin transactions were related to crime”; now it is “less than 1% — a lower percentage than for fiat.
“Every time you speak to someone in law enforcement, they’re always happier to trace crypto.” Tracing is not recovery, though, funds traced to somewhere like North Korea are hard to retrieve, unlike money traced to a UK bank, Stanford notes.
AI lowers the barrier to fraud
AI has made access to fraud cheaper and easier too, particularly in relation to impersonation, where “videos or audio that look and sound exactly like the person they’re imitating”, have proliferated.
A crime-as-a-service market sells scams-as-a-service and money-laundering-as-a-service tools, so “nobody really needs any technical skill … to scam,” says Stanford.
Deepfakes and AI-written messaging have “removed a lot of the ways you might traditionally have known that something is a scam.” Criminals stay ahead “because they’re not limited by things like law or data protection.”
Banks balance protecting customers against locking them out of their own money. Stanford says friction is often enough: “Scams and frauds rely on pressure and urgency … sometimes all that’s needed is adding a delay.” Liability should not always fall on the bank, she added, citing customers who send money to scams “knowing that they will get reimbursed.”
She linked the UK’s APP fraud reimbursement rules — a world-first requirement from the Payment Systems Regulator — to a rise in UK fraud, calling it “a direct correlation,” though the rules remain contested. Social media, advertising and messaging platforms should carry more responsibility: fraudulent ads often stay live, and “there just isn’t enough done between a scam being reported and it being blocked.”
What firms should do
Asked for the one change a payments firm should make in the next year, Stanford said education: fraud and the networks behind it “are learning and developing faster than legitimate companies can keep up.”
Many people miss basic risks, like an unknown link, a spoofed sender, or a recruiter’s PDF that “maybe … is a North Korean hacker trying to get you to download malware.” Her advice: “Educate people to be more cautious and to not trust anything inherently.”
The latest episode of the Payment Expert Podcast is available now on Spotify, YouTube, Apple Podcasts, and all major podcast platforms. Subscribe to the Payment Expert newsletter so you don’t miss an episode.
