The Download

Illustration: Morning Brew Inc., Photos: Adobe Stock

TL;DR: A growing number of AI agents and tools are now helping you shop until you drop—from recommending products to completing purchases in one click. It’s a bet AI companies and brands hope will pay off big time—but it could also worsen your impulse buying habit.

What happened: Yesterday, TikTok launched an AI shopping assistant and one-click checkout, making it even easier to snap up that viral magnesium supplement on your For You page. These features arrive weeks after Meta’s splashy release of Muse—with one of its key selling points being its ability to shop for you. Other AI companies are rolling out shopping aids, too, like ChatGPT’s new virtual “try on” tool for clothing and accessories.

What makes TikTok’s and Meta’s shopping push especially significant though, is that they already drive a lot of online shopping. Meta is an advertising juggernaut (take Instagram’s eerily specific ads, for example) and runs Marketplace, while 54 million Americans now say they’ve bought something after seeing it in a TikTok video.

Easy to shop…and spend: Agentic shopping’s big draw is that AI can cut the time it takes to find the perfect slow cooker or pair of jeans. But all that convenience also has a potential cost: Swiping your credit card becomes a little too effortless.

Some early adopters say they’re already shopping way more than they did without a robot assistant. And Instinct, another popular AI agent, apparently keeps sending users product recs they didn’t ask for—in one case, it reportedly bought an unneeded bikini without the user’s approval.

Following the money: The agentic shopping spree comes as AI firms face two realities: Running an AI lab is expensive, but few consumers are paying for the tech—only 2.2% of US households had paid AI subscriptions as of May. (OpenAI, for one, is now trying out other ways to boost revenue, like serving ads alongside images its models generate.)

That might be why Meta is leaning so hard into shopping with Muse—since the company expects most users to stick with the free version. It’s angling to eventually profit by taking a small fee on each Muse transaction, but that only works if people get comfortable handing an AI agent their credit card—and let it talk them into buying the $80 sweater they don’t really need.

Bottom line: If AI shopping takes off, it could become a vivid example of why friction is sometimes good—especially when it comes to spending money. —WK

      

Sponsored By Chargebee

Flat per-seat pricing is great. But advances in AI disrupted the fundamental unit economics that made it make sense. As a result, new pricing metrics are popping up everywhere (think pricing active agent minutes, gaming-like prepaid credits, usage-linked seat tiers).

Most billing systems weren’t built with these in mind. Thankfully, that’s where Chargebee can help. They build the metering, billing, and monetization infrastructure that helps you scale new pricing models with more agility.

So instead of asking your in-house team to launch a multi-engineer, multi-quarter project, you can use one single billing system that accommodates any model.

Give your team the freedom to experiment with new pricing. Start a free trial of Chargebee.

Stop screenshotting your videos

You’re debriefing your friends on last night’s concert, where you filmed every song with your iPhone like it was your job (unless the performer was Phoebe Bridgers, in which case we can’t help you). But you also don’t want to subject them to that two-minute clip where you’re screaming your head off.

Until now, saving an image from a video required taking a screenshot, which meant a weird aspect ratio and an often blurry downgrade in quality. Good news: A new feature in iOS 27 finally lets you pull a photo straight from the clip.

How to save a photo from a video: Open the Photos app, then tap the video. Drag along the timeline and pause on the frame you want. Tap the three dots in the top right corner → Save Video Frame as Photo.

The photo saves to your library at the same resolution as the video—and since iPhones shoot in up to 4K, that’s plenty of detail. It works on any iPhone running iOS 27 (iPhone 11 and up). —LC

If you have a tech tip or life hack you just can’t live without, fill out this form and you may see it featured in a future edition.

The Zeitbyte

Rage against the (new) machine

Illustration: Shannon May, Photos: Adobe Stock

In a world of yearly iPhone releases and new AI-powered devices announced seemingly every day, society’s strongest soldiers might be a small but committed group of “upgrade holdouts”—people who steadfastly refuse to replace their old tech, per The Atlantic.

Take the 66-year-old nurse who still blasts music on his 2003 iPod. Or a 54-year-old consultant who still makes calls on an iPhone 5s, a device unveiled over 13 years ago. (He also uses a pair of headphones that he got from RadioShack in 1983—the year Thriller topped the charts.) And one of the people The Atlantic interviewed uses an old-school flip phone, borrowing his wife’s smartphone when he encounters a QR code.

Some of these old tech loyalists are fed up with what they suspect is tech firms’ planned obsolescence (and they doggedly fix up gadgets using online guides). Others just don’t want to learn how to navigate a new interface. But even the most determined upgrade resisters know that they’ll have to yield one day: The iPhone 5s user, who lives in Canada, might see his device become a glorified calculator next year as carriers shut down their 3G networks. The bleeding-edge iPhone 6s awaits. —WK

Chaos Brewing Meter: ☕/5

• ChatGPT is now rolling out its invisible text watermark—and your workplace can opt in too.
• Adding money to Apple Cash will soon require one extra step.
• How much did Anthropic CEO Dario Amodei make last year? Less than you may think.
• Can you tell a real New Yorker cartoon from an AI knockoff? Hint: The signatures won’t help you.
• Google Docs and Drive just got a pretty handy upgrade that plays nicer with AI.

^{*A message from our sponsor.}

This time last week...

Readers’ most-clicked story was about the humans reviewing Microsoft Copilot prompts. They say they’re drowning in disturbing image requests.