Snippings

xxx

While Fiserv said it won’t disclose customers’ names to one another, the fact that it’s training the chatbot on enterprise customer data could be of interest to countless companies that would love to do the same.

 

From: $120 Billion Fintech Firm Trains Chatbot—on Customer Data — The Information.

xxx

 

Fiserv did not respond to further questions about how it managed to obtain permission from those customers to use their data for chatbot training.

The United Nations has proposed a mandatory digital ID system linked to bank and mobile money accounts to enhance social protection programs while addressing online misinformation.

North Korean hackers have successfully cashed out the first $300 million of the proceeds from their record-breaking $1.5 billion crypto theft from the ByBit Exchange. They are getting good at this: crooks linked to North Korea have stolen more than $6 billion in cryptoassets since 2017 (with the proceeds reportedly spent on the country’s ballistic missile program). Elliptic says the incident is almost certainly the single largest known theft of any kind in all time, a record previously held by Saddam Hussein, who stole $1 billion from the Iraqi Central Bank on the eve of the 2003 Iraq War.

That’s the thing about anonymous, untraceable digital assets: the bad guys have them too. This incident nudges the crypto community between the Scylla of crime and Charybdis of censorship. Do you accept anonymity, and accept the societal harms that results on the basis that overall it is better for citizens going about their everyday lives to avoid surveillance or do you accept censorship and accept that not all transactions should be allowed?

xxx

Perhaps surprisingly, most check payments were made by businesses in 2021, 52 percent by number and 76 percent by value. This is the first year for which the FRPS found that the number of business checks exceeded the number of consumer checks. In addition, check use by value increased from 2018 to 2021; the total value of business checks increased by 2.7 percent per year to $20.64 trillion.

From: Business Use of Wires Grew Sharply from 2018 to 2021 – Federal Reserve Bank of Atlanta.

xxx

xxx

Our latest report shines a light on what we found, and key highlights include:

Fraudsters target major platforms to commit scams: Social media, technology platforms, and telecoms are frequently exploited by scammers, resulting in losses of hundreds of millions of pounds for UK consumers. 
Over half of scams involve Meta platforms: In 2023, Meta platforms (Facebook, Instagram, WhatsApp) were linked to 54% of scam incidents (119,338 cases) and 18% of total losses (£62.7 million). That’s roughly £1 in every £5 lost in scams. 
Telecoms and emails cause significant losses: Fraudulent calls and texts via telecoms were used in 12% of scam cases (26,975) but accounted for 31.5% of the losses (£107.2 million). Emails, while only involved in 2% of scams, led to disproportionately high losses of £35 million (10% of the total). 
Romance scams are more common on Meta platforms: Meta platforms were used in more romance scams (31%) than all dating websites combined. Facebook, Instagram, and WhatsApp together accounted for 1,590 incidents and £5.1 million in losses. 
Purchase scams dominate on Facebook: The most common scams in 2023 were purchase scams, where victims buy items that never arrive. Facebook was used in 44% of these cases (67,337 incidents), with losses of £19.5 million—more than any other platform. While eBay saw fewer scams (1.6% of cases), the losses per scam were significant, totalling £6.7 million. 
Investment scams cause the biggest losses: Investment scams made up 23% of total losses (£80.3 million) but only 6% of scams (12,500 incidents). Fraudsters often used telecoms (£18.4 million in losses), Meta platforms (£11.6 million), and even friends and family connections (£9.6 million) to deceive victims.

From: Unmasking how fraudsters target UK consumers in the digital age | Payment Systems Regulator.

xxx

xxx

Sony Music has revealed the scale of its battle with artificial intelligence fakes of its artists by saying it has taken down more than 75,000 examples of AI-generated material featuring its biggest stars, including Harry Styles.

From: Sony Music says over 75,000 items removed in battle against AI deepfakes.

This is a Red Queen’s race. The industry should be focusing on labelling real content, not trying to find illegal content (of which there will be an effectively unlimtied amount).

Andrew Tobin, a thorough scholar of these issues is optimic. He says that as a digital identity for legal persons (including businesses) is coming as part of eIDAS 2.0 so what he calls a “LPID” (legal person identity data) credential is already on the cards, to be stored in an organisation’s digital wallet.

xxx

Let’s assume that our personal AIs will become smart enough to handle the things people don’t want to do. So what’s left is… the real customer engagement. Those steps, those moments, where customers want to be engaged.

In other words, by automating and eradicating the Bad Stuff, we can strip things back to only the Important and Engaging Stuff.

Let’s call it Good Friction.

Service designers already know this. When something is too easy to do, too frictionless, it loses its appeal. Customers get suspicious.

From: The unexpected opportunity with AI agents – designing for customer friction | LinkedIn.

xxx

xxx

Two students at Harvard University have hooked Meta’s Ray-Ban smart glasses up to a facial recognition system that instantly identifies strangers in public, finds their personal information and can be used to approach them and gain their trust. They call it I-XRAY and have demonstrated its concerning power to get phone numbers, addresses and even social security numbers in live tests.

The system is perfect for scammers, because it detects information about people that strangers would have no ordinary means of knowing, like their work and volunteer affiliations, that the students then used to engage subjects in conversation.

From: Meta’s Ray-Ban Smart Glasses Used To Dox Strangers In Public, Thanks To AI And Facial Recognition.

xxx

xxx

The European Central Bank (ECB) announced it’s expanding the settlement of transactions recorded on distributed ledger technology (DLT) in central bank money with a two-track approach.

First, as soon as feasible, the Eurosystem will develop and implement a safe and efficient platform for such settlements in central bank money through an interoperability link with TARGET Services. No timeline was provided.

From: Eurosystem to connect DLT settlement with TARGET – Finadium.

xxx