When you drop a new Substack

Stripe’s Link wallet is fast becoming the de facto payment option for many personal agent experiences — and interestingly they offer a degree of purchase protection. (US only for now.) The terms of the Purchase Protection were recently updated and essentially works like a purchase insurance rather than as a payment scheme chargeback. Cover is provided by New Hampshire Insurance Company and Cover Genius with Stripe providing the service as “a convenience and benefit”, in other words it’s not a legal, or scheme obligation, and could be withdrawn in the future. Any claims are administered by Cover Genius and not Stripe. This is a different model to what most people are used to with purchase protection, in which queries usually go via a card issuer. As agentic commerce grows in importance will we see third party insurance models grow in popularity? It makes sense given many of the rules around liability are not always 100% clear, and traditional routes such as chargebacks are still adapting. Also noting again that this is US only for now. The US market has much higher interchange than the UK and EU, therefore some of the transaction revenue could essentially be used to pay for an insurance model of protection — this will be much harder in the UK and EU. onelink.com/gb/protections

What I like about Instinct is it’s both a company name and a product name Duet Dots Names, which are products of a wider portfolio of big AI labs don’t hit the mark in the same way Muse is better… It’s easy to say “My Muse” and they created a character distinctive enough than you can forget it’s Meta

This year I stopped my @TheEconomist subscription after 20 years @WorksInProgMag is replacing it for my in terms of reading time The format of a weekly newspaper is tough to pull off in a world where news moves so quickly — a world of Substack, notifications pinging and daily updates It wouldn’t surprise me if The Economist eventually moves to more long form reports and content and less news over time

I'm starting a fintech and payments podcast. The goal is interesting discussions rather than high production quality. Remote recording (self produced). Looking to talk with those building interesting stuff and get across product, tech, business, and strategy. Please no PR pitches but open to any recommendations or direct outreach.

For a politics podcast nerd like me this was a fun day at #battlefest Thanks to @LiamHalligan, @AllisonPearson , @Steven_Swinford, @patrickkmaguire, @Scarlett__Mag, @JAHeale, @lara_e_brown And many others I saw today

In its latest financial results, Revolut revealed that revenue from paid accounts reached £708 million, an increase of 67% year-on-year, while paid subscriber numbers grew by 42%. Revolut doesn’t disclose how many paid subscribers it has, but estimates suggest around 7.4 million subscribers, assuming an average of £8 per user per month during 2025. The UK's other major neobank, Monzo, has >1.6 million paid subscribers, likely all in the UK at this stage. Both companies are enhancing their paid accounts and also increasing their prices. Revolut's Metal account is getting: - A price increase from £14.99 to £16.99 per month. - The addition of five new subscriptions including £10 monthly Laundryheap credit, a Loveable subscription, and Super Duolingo. - Additionally, Metal subscribers will get an extra 2GB of roaming eSIM data. - These additions are on top of existing benefits such as a Financial Times digital subscription. Monzo's Perks account is getting: - A price increase from £7 to £9 a month. - The addition of YouTube Premium Lite, which offers ad free viewing on most videos except notably YouTube music (usual cost £7.99). - Existing benefits include a free Vue cinema ticket each month, a weekly treat from Greggs, and an annual railcard for discounted train travel in the UK. Whether these accounts are worth it really depends on your personal circumstances and if the benefits make sense to you. Overall it's great to see this competition for paid subscribers. For many years the paid current account market saw little differentiation, and that's all changed with the neobanks.

For AI to be a true General Purpose Technology (GPT) it needs a better way of interfacing with consumers day-to-day. Until now, for many consumers AI has meant Chat GPT (more or less), or something similar. Actual Agentic Commerce stalled… And saying this would be a kind way of putting it. Despite the technical interfaces, protocols, and announcements, there was almost no way for consumers to participate in Agentic Commerce. Technology accelerates when interfaces are easy to use and accessible — not fragmented and specialised. Until recently Agentic Commerce meant an ever growing protocol soup of acronyms. Personal Agents change this. They give a “normal user” ways to interact and buy with agents via an interface(s) used every day. This can’t be overstated. Until now consumers were not interested in Agentic a commerce, because essentially as they couldn’t use it on an end-to-end basis. They couldn't experience it. A lot of technology doesn't make sense until you use it for yourself. Personal Agents open up full Agentic Commerce from discovery to checkout to anyone. That doesn’t mean to say Personal Agents solve everything, but there is a strong case that they have, finally provided us with the "missing link" in paying with agents — a full year since Stripe and Open AI first announced Instant Checkout.

I wrote about this here paymentsculture.com/p/the-ge…