The esteemed Simon Taylor nails it with Apple and AI in payment space. An insightful read…

Written by Michael E Dehn

Founder and CEO of Metro Pulse a continually running enterprise since May 1980.

June 16, 2024

What if Apple Intelligence worked with Open Finance?
Apple has stuck the landing on AI. The launch of Apple Intelligence has excited people, and despite its being “boring” in features, it looks usable. This is classic Apple. Apple has built UX and privacy as a moat. Whenever Apple sticks the landing on a technology, it reshapes the market. While current features are limited, what excites me is what could happen with AI + the Apple Wallet.
Apple Wallet is the natural home for a private, secure, consumer-advocate AI that helps co-pilot or even becomes an agent to manage your finances.
What is Apple Intelligence?
Users can now get writing help, AI summaries, simpler notifications, better photo editing, and Siri, which is much better at handling imperfect speech. It is not a fully-fledged ChatGPT or Claude equivalent. However, it is incredibly restrained and helpful for users without the risk of hallucination or the inconsistency of interacting directly with the foundation models.
This means users can find photos with searches like “beach photos of a girl,” and it will return useful results and remove unwanted people from the background. It can generate images or respond via Siri in very limited context. But if you want to generate a new story or an image type they don’t support, Apple will push you to ChatGPT instead.
This is smart, given the Gemini controversy – but the most interesting thing is precisely how Apple sends information to ChatGPT under the hood.
Apple’s Privacy Moat gives them brand permission to handle sensitive data like finance.
Apple has made a name for itself as the privacy champion. It makes more money from selling devices than it does from data, so there’s a lot of upside to that marketing message. But there’s also a legitimate case, with end-to-end encryption as a default across most services.
Apple has solved some genuinely hard engineering problems to make AI truly private.
It’s enabled AI in a tiny 3 billion parameter model that lives on the device’s secure enclave.Only if the device can’t complete a task locally, it sends the request to a model in an Apple “Private Compute” cloud. While most cloud storage is encrypted, data has to be decrypted in memory (and being used). Apple’s cloud replicates the on-device secure enclave to ensure data is never unencrypted in memory or during network transport.The phone defines which model is needed, packages your prompt, and locks your data; only the corresponding model can “unlock” that data. (i.e., PKI).From there, the private compute cloud immediately deletes any user’s data.The entire process is open to review by independent researchers (which is not true of many other foundational models and providers).
This privacy moat extends to financial services. The secure element and enclave was critical to putting cards in the Apple Wallet. Today, many consider Apple Wallet much more secure than traditional payment methods and lower in fraud or chargebacks (although that’s changing).
The Wallet is the Future of Fintech, and the Apple Wallet is dominant in the US.
The Wallet is the center of gravity for consumer Fintech. A decade ago, card issuers wanted to be “top of wallet.” Today, the battle is to be the Wallet. Consumer identity, tickets, cards, BNPL, and other payment rails are all live in the Wallet.
Data from Pymnts.com says 12% of consumers use mobile wallets as their primary payment method, half (6% of all) use Apple Wallet, and 4% use PayPal/Venmo. Apple continues to see adoption rise, and now it supports more P2P features; I imagine that will continue.
As I wrote, “Apple will be a bank in all but name.” Apple offers almost every financial product.
Gen AI is coming to consumer Fintech
A quick scan of recent 4 Fintech Companies shows a pattern:
Korzo is an SEC-registered investment advisor who describes itself as “your money buddy.”Alina is a chatbot and AI agent for investing aimed to get women investingXillion will analyse your consumer portfolio from screenshots or PDFs to help you optimize
ChatGPT isn’t bad at handling transaction data if sufficiently well prompted.
But the risks are massive.
As the OCC recently outlined, the risks of AI from fraud, misinformation, and cyber attack are nontrivial. “Prompt injection” is the vulnerability that exists when you take a carefully crafted prompt and add additional input that changes the meaning.
And there is no way to stop it at the foundation model.
Given the privacy concerns and the unpredictability of the foundation models, the consumer need for a well-packaged experience is obvious.
What’s less obvious is if this is a dedicated consumer app, a feature in an existing one, or if it all ends up back in the Wallet.
Apple has a habit of “borrowing” ideas from apps, and this will happen with Fintech apps.
If an app does well in the App Store, a few years later Apple has a habit of releasing its own version of that App. Everything from health apps, display settings and even maps have been “Sherlocked.”
Sherlocking is when Apple introduces a new feature that renders a third-party tool irrelevant.
This year, Apple introduced apps that kill Password Management (e.g., LastPass), Voice transcription (e.g., Otter), and Phone mirroring (e.g., Bezel). Now, you could argue these services will a) Continue to exist b) live naturally in the OS c) Google, Amazon, and Microsoft all do this.
The point is. This will come to Fintech, and specifically Apple Wallet.
Apple Wallet is the natural home for a private, secure, consumer-advocate AI that helps co-pilot or even becomes an agent to manage your finances.
What if Apple Intelligence + Apple Wallet?
Imagine getting insights and recommendations for every card, account, or loyalty card in your Wallet. Apple could help you select the card that maximizes rewards at checkout, set up a savings habit, and even ensure you’re maximizing your tax advantages with brokerage and savings.
With Siri, you could “talk to your money” and get coaching on improving. The ability to handle pictures and documents means Apple could make sense of your taxes in the context of your transactions.
Of course, Apple could do many of these things quickly. But it won’t be ready for the market until it’s very sure the UX and privacy are so robust.
Open Finance + Apple Wallet + Apple Intelligence
Apple recently launched open banking in the UK to show transaction feeds in the Apple Wallet. With the passing of Dodd-Frank 1033, my mind immediately jumped to the idea that it would open the door for more in the US.
It’s one thing to make sense of everything in the Apple Wallet, it’s another to be able to make it the hub for taxes, brokerage, mortgages, auto loans, and much more.
While 1033’s initial scope is limited to consumer checking, savings, and lending products, it’s not fantasy to suggest that it can and will expand through partnerships with data aggregators. Apple acquired a UK account aggregator (Credit Kudos); would it do the same in the US?
A glimmer of the future of Fintech
The future of Fintech belongs to those who master the convergence of privacy, AI, and secure computing at the device level.
Their Apple Wallet already demonstrates this as a smart vault for payments, passes, keys, and ID that’s always secure and current. But Wallet is just the beachhead.
Their secure enclave, Wallet, and now moves in AI could unlock a lot more in the future.
Of course, we should take all of this with a grain of salt.
Apple moves in tiny increments.
Much like their device engineering, these increments are often so small that you barely notice them these days. In fact, when they do take a big swing (like Vision Pro), they often miss the mark.
But don’t confuse that for a lack of action.
Every release is another brick in a giant wall.
Year by year
The Wallet becomes dominant.The secure enclave becomes more useful.More payment types are added (like BNPL).More official Government identities are added (like drivers licences).
And now we have AI.
Consumer financial services is actively being commoditized by Apple.
As an avid Pixel Phone user, it annoys me that Google had all of this in their grasp with Google Pay, or Google Wallet, or whatever it’s called this week.
Apple will undoubtedly be a huge player, but not the only player. The fact that they can do something doesn’t mean there will be no competition.
What I love about looking into the future is working back to today to figure out what we should do next.
What will you do?
ST.
 
4 Fintech Companies 💸
1. Fence Finance – Credit facility automation
Fence finance helps credit funds and lenders get real-time funding for every loan they originate. Lenders don’t have to pay the fees for creating a warehouse debt facility, and credit funds avoid having to manage capital calls. Fence finance helps automate much of the due diligence a credit fund would have to do and then issues the capital in real time if the lender passes.
🧠 This is like merchant cash advances but for private credit and all real-time settled. Imagine you’re a BNPL lender. Today, you might have a $200m credit facility with a bank, but of that, you might receive $1m every 15 days. If there’s a spike in loan activity, you might rely on a warehouse (a big loan), but that’s costing you in APR. The fund or bank you receive it from will occasionally have to check to ensure your lending still fits inside the agreement, which both parties can do manually. Automating all of that and making it all real-time is a big deal.
2. Frich – Anonymous wealth benchmarking for under 25s
Frich helps younger consumers understand how they compare to peers anonymously via an app. Users can compare salaries, spending habits, and rent payments. Frich asks users to submit their most uncomfortable questions and win $100 if its chosen.
🧠 This is a great twist on “financial education.” Because it’s leaning into the questions everyone has but is afraid to ask on their terms. What’s clever here is that all users submitted and permissioned it. I imagine there are plenty of businesses that would want a similar capability. They aim to use this community and data to help target ads and financial products to users over time.
3. Daloopla – The equity analyst’s AI analyst.
Daloopa helps update spreadsheet models for public stocks using real-time market data. It plugs actual numbers into Excel models and helps quickly identify new guidance and GAAP adjustments. Users can model an entire industry and update it with actuals.
🧠 Two themes stand out. 1) AI is giving spreadsheets superpowers. 2) Capital markets x AI keeps happening. It’s augmenting its AI with humans in the loop to ensure data quality. This is solving having 50 tabs open to copy and paste data from various PDFs or data sources. Hidden schlep work begone!
4. Klearly – The Mobile only Point of Sale
Kearly provides a simple payment acceptance solution for small and micro merchants in the Netherlands. Instead of a taxi driver or small store having a terminal, they can enter the amount on the merchant phone and have the user tap their card or phone to authorize a card payment. Kearly claims to be consistently cheaper than its competitors.
🧠 Someone had to differentiate on this as a wedge. Since mobile phones have become a point of sale, they will inevitably become a feature. But making it the star feature and the wedge proposition makes sense.
 
Things to know 👀
1. Affirm BNPL added to Apple Pay
BNPL with Affirm is coming to Apple Pay. Users will be able to “Tap-to-BNPL” with Affirm. Apple also said installments will be available with cards from Citigroup, Synchrony, and Fiserv-related issuers.
🧠 Apple already had Pay Later, so why partner? My guess is Affirm’s underwriting model is proven. Apple is learning what bits of the finance stack it wants to own and doesn’t.
🧠 Wallets are the future of payments. Every payment type you see at e-commerce checkout will slowly appear inside Apple Wallet.
🧠 BNPL is a mainstream payment option. For all the negative press and FUD from the bank lobby about BNPL, it is consistently proving its viability.
🧠 Remember Mastercard installments? It’s interesting that the card networks haven’t played well in BNPL. They’re better placed to enable wallets and new payment types. It seems Mastercard is headed that way now (e.g. by tokenizing all e-commerce transactions in Europe by 2030).
2. Apple launches tap to cash
With Apple Tap-to-Cash, users can send and receive Apple Cash by holding two iPhones together. They can also redeem rewards or manage installments through their eligible debit or credit cards directly from their wallets. Ticketing also got a boost with stadium details and recommended playlists for upcoming events.
🧠 When will we see Apple Intelligence driving Apple Pay? Apple has put in place the pieces of private data and compute. If anyone can start getting closer to finance with AI, they’re in the primary position.
🧠 What happens if you combine open banking (1033) and Apple Intelligence? Imagine if my device, with its private, secure AI, could help be my personal CFA or CFO.
🧠 The future of Fintech is not between cards or banks; it’s between wallets. A wallet can hold your identity and all of your financial apps and manage ticketing.
🧠 Slowly, then suddenly, the Apple Wallet is becoming THE Wallet. Apple is in no rush but is in the lead. Pymnts reported that of the 12% of consumers who use wallets more than any other payment, half use Apple Wallet. PayPal is losing market share YoY.
 
Good Reads 📚
1. OCC Speach about AI and a Shared responsibility model.
The nature of how AI “learns” diffuses responsibility. Like all technologies, there are risks, but it’s unclear who should manage them. The OCC Comptroller proposes a “shared responsibility model” to help manage fraud, scams, and ransomware attacks.
The comptroller also uses a model from the shift to electronic trading as a metaphor, how we moved from tech helping with inputs for traders to consider, co-pilots for traders to do more faster, and finally, agents who execute trades automatically. AI is doing this for many more task groups. The regulator notes concerns in the following order
The volume of fraud due to deepfakes, voice cloning, sophisticated heists, and social engineering.AI-enabled cyber attacks from nation-states aimed at attacking financial infrastructureAI-generated disinformation (like a Russian AI-generated “attack on the Pentagon” that went viral)
How do we ensure AI is safer from these attacks? How do we ensure AI is accountable (e.g. if a Chatbot refuses to offer a refund, is that OpenAI’s issue or the bank?) And how do we ensure that AI is fair when used for underwriting and not discriminating? The comptroller proposes a model similar to the one used in cloud computing.
🧠 Do we need a FINRA for AI in Finance? Or an industry association like NACHA or The Clearing House? It’s a great question the comptroller raises. I think the answer is Yes. But what if it was an SRO or industry association that could focus on novel activities broadly. Such as the use of AI, Banking as a Service, etc?
🧠 This SRO or industry association would define clear responsibilities and risk approaches. This is similar to how NACHA, The Clearing House, and FINRA have rulebooks and guidelines.
🧠 It’s an elegant solution because it’s not a new agency or regulator. It doesn’t require an act of congress. It’s something the industry can lead and self-select into. You would auto-select into this if you care about users, trust, and credibility.
🧠 The hardest part with this stuff is the incentives issue. Often, the big banks have the deepest pockets, so they end up footing the bill, and we need to find a way to make access equal. The best funder for something like this would be a mix of benevolent (e.g., Omidyar), industry (banks, Fintech companies, VCs), and, if possible, a contribution from USG.
 
🕊 Tweet of the Week
Jason Mikula@mikulajaOk, I’m in the airport lounge and actually reading the Evolve cease & desist.Here’s what catches my eye:1. Safety & soundness examination of EBT’s “open banking division” (BaaS) in Aug ’23 identified risk management & consumer compliance deficienciescon’t… Jun 14, 2024  77 Likes   9 Retweets   5 Replies
That’s all, folks. 👋
Remember, if you’re enThat’s this content, please do tellyou’reour fintech friends to check it out and hit the subscribe button 🙂
(1) All content and views expressed here are the authors’ personal opinions and do not reflecauthors’ews of any of their employers or employees.
(2) All companies or assets mentioned by the author in which the author has a personal and/or financial interest are denoted with a *. None of the above constitutes investment advice, and you should seek independent advice before making any investment decisions.
(3) Any companies mentioned are top of mind and used for illustrative purposes only.
(4) A team of researchers has not rigorously fact-checked this. Please don’t take it as gospel—strong opinions weakdon’tld
(5) Citations may be missing, and I’ve done my best to cite, but I will alwaysI’ve to update and correct the live version where possible. If I cited you and got the referencing wrong, please reach out
 
Update your email preferences or unsubscribe here© 2024 Simon Taylor.1 Royal Exchange
London, Greater London EC3V 3DG, United Kingdom of Great Britain and Northern Ireland

You May Also Like…

Hmmm

https://twitter.com/PaulIngrassia/status/1815124668564615337

YAY!!

https://twitter.com/LahavHarkov/status/1815082831007174659