On Nov 18 my AI agent found 10% off at HOKA with a free membership, plus cashback, even though HOKA's site blocked its browser.
The offer came to the agent directly, so nothing had to be scraped.
If your agent can't see the store, how does it know you got the best deal?
1
38
Klarna's shopping browser, for price tracking and cashback, has driven $34B+ in purchases since 2020.
Klarna is now testing it in an AI assistant on WhatsApp and iMessage with early users. The cashback check is moving into chat. At Minty we build that layer for shopping agents.
3
84
@Klarna Q3 release, Nov 17: browser "has driven more than $34 billion of GMV since its launch in 2020"; AI assistant "live with a small group of early users, with pre-registration open, on WhatsApp, iMessage and in the Klarna app." s205.q4cdn.com/644747736/fil…
41
If an AI agent shops for you, the cashback should still be yours.
Riskified can now tie an AI agent's order to the real shopper. Once stores know who's behind the agent, the reward can follow the person. We're building Minty as the cashback and offers layer inside those agents.
2
1
99
Riskified Q3 release, Nov 11, 2026 (revenue +41% YoY to $115.2M):
businesswire.com/news/home/2…
"...identify the AI agent behind an order or customer service request, resolve it to a real consumer identity, and return a risk signal in real time."
1
47
One AI said it couldn't see which card a Best Buy checkout would charge.
It told the user to check the last 4 digits on the final screen. That's the step I'd copy. Card benefits follow the card that's charged.
2
117
I'm paying for [item, $] at [store or app]. My cards: [list].
1) Which card will this checkout charge?
2) Each card's purchase protection and extended warranty.
3) What each earns here, in $.
Pick one, say why, flag what you can't confirm.
Chase, Oct 29: creditcards.chase.com/cash-b…
1
1
82
Honestly the most interesting part of Affirm's letter wasn't the numbers. Max says your AI agent won't let you miss the 0% APR deal. That's the agent finding offers, not just checking out.
1
80
Max Levchin, p.4: "Your favorite agent will delight you with the perfect gift idea for this holiday’s shopping list… and it most certainly won’t let you miss out on the sweet 0% APR offer from Affirm."
1
50
Source: Affirm FQ1'27 Shareholder Letter (8-K Ex 99.1, filed Nov 5, 2026), p.4. sec.gov/Archives/edgar/data/…
9
On Apple's own site, one Apple Watch costs $50 less certified refurbished than new, and the card wasn't the big number.
That's why the prompt in the image asks for the certified refurbished price first, then the card. Exact prices are in the reply.
1
57
Prices from Apple + Citi pages Nov 5 ~3:31 PM ET, pre-tax.
Apple Watch SE 3 GPS+Cellular 44mm: $329.00 new, $279.00 Apple certified refurbished. Flat 2% card (Citi Double Cash) $5.58.
2% on new too: $49.00 gap.
Example, not financial advice.
apple.com/shop/refurbished/w…
39
Perplexity Teammates are Max-only on the web for now, with mobile later.
The interesting part isn't that they run Computer tasks. It's that they can message coworkers in Slack and coordinate with other teammates without waiting for another prompt. The hard problem moves from "did the agent answer right" to "who authorized the outbound message."
PERPLEXITY 🔥: Teammates, Perplexity's new proactive agents, are rolling out to Perplexity Max subscribers on the web!
Teammates are running in an isolated environment and can execute Computer tasks, skills, memory, and tools.
Coming soon to mobile apps, too!
Testing time! 👀
55
The trust fixes that exist on paper are about the card. Shopify's Q3 call: the agent never sees your card details because Shop Pay is the wallet. Stripe says Muse asks you to approve the total on every purchase. An agent can still do something embarrassing with a perfectly protected card, and that's the failure type this episode calls scariest.
David Pawlan, creator of Assistant Benchmark, has tracked 147 AI assistants so far, and in his group chats of AI power users, the number one fear is trust. It comes up in 22% of roughly 3,600 messages.
He puts agents through head-to-head tests, like ordering same-day flowers as cheaply as possible or planning a weekend in LA that's both luxurious and cost-efficient, to see which ones can hold competing preferences.
He joins a16z's Anish Acharya and Josh Elman to discuss why the scariest agent failures are the embarrassing ones, why agents are where e-commerce was in 1997, why scale is still a moat, and the "dark agent" that could speak to the parts of ourselves we don't reveal.
0:52 Inside a meetup for 100 agent nerds
2:52 Why we'll all have an army of agents
5:36 Trust is power users' top fear
6:28 Building fences around your agent
10:14 Privacy loses to free pizza
12:41 Can agents balance price and speed?
14:19 How agents define luxury differently
19:44 Migrating from one agent to another
21:06 What your photo roll tells an agent
25:48 Will agents make us more rational?
27:38 Why paternalism kills products
31:43 Gossipy agents as social glue
33:41 Scale is still a software moat
36:55 From OpenClaw to agents hanging out
39:23 Agents that introduce you to people
46:15 Predictions and the dark agent idea
YouTube: youtu.be/dSiTwPl0t1U
@DavidPawlan @illscience @joshelman
1
50









