Plus, you can now lease an iPhone directly from Apple.
July 28, 2026View Online | Sign Up | Shop
Newsletter Logo

Presented By

Sponsor Logo: PwC

Good boy or naughty photo? Apparently, iMessage’s Sensitive Content Warning can’t always tell. The feature flagged a video of a dog getting belly rubs as nudity, according to a 404 Media report. And this wasn’t a one-off case: Apple forums show similar reports of false flags on animal pics over the past few years. 404 notes that Apple hadn’t responded to a request for comment. Presumably, it’s still running a background check on the dog.

Also in today’s newsletter:

  • Google settings you should turn off now.
  • A Sunday service for future AI founders.
  • Apple’s leasing program goes live.

—Lindsey Choo, Whizy Kim, and Saira Mueller

The Download

Now available to rent: Your face

Photo illustration showing 3 browser windows layered on top of each other, each one with a close-up of a different person's face. There's a button at the bottom left with a face icon that says

Shannon May, Photos: Unsplash

TL;DR: In China, ordinary people are licensing their faces on likeness catalog platforms for use in AI-generated content, typically making between $15 to $700 (sometimes per use). It could be a glimpse of a future where your face is monetized and treated like IP.

What happened: Anyone who wants to “rent” a face can browse by categories like age, gender, or even “type of face” (like “rugged” or “supermodel”), per Rest of World. As a licensee, you set prices based on different factors and how many uses you’re permitting.

The hot use case is for microdramas: the vertical video serials (with episodes typically one to two minutes long) that have taken China by storm—though they’re gaining ground in the US too. Over 128,000 microdramas were released in China in the first quarter of 2026. A massive 95% of them use AI, probably because it’s cheaper than hiring actors. And basing the AI clones on real faces means less homogenous-looking characters (something viewers on social media have apparently been grousing about).

The appeal for people renting out their faces: AI content firms are stealing people’s likenesses anyway (and not just those of famous celebrities)—so you might as well have some control over it and make some money in the process. In the city of Guangzhou, a court that deals with cyber-related lawsuits has heard about 700 cases of AI face theft in the last few years, according to a report from a national newspaper in China.

Face snatching gone global: Microdrama actors in the US also claim they’re seeing AI versions of their likeness appear in shows they never actually acted in. These roles are often nonunion jobs that give “broad editing permissions” to productions, according to an actress interviewed by Business Insider. Others have licensed their faces and voices to TikTok’s avatar tool for around $500 to $1,000 dollars (and regretted it.)

Is that even legal?: Many states have some kind of protection against using your likeness for commercial use without permission—and celebs like Matthew McConaughey and Taylor Swift have filed trademarks on their images and voices (all of McConaughey’s were granted). But it could be tough to prove an AI character stole your face, instead of it just being an uncanny coincidence.

Bottom line: AI makes it too easy to clone anyone’s appearance, whether it’s a top movie star or just your neighbor who posts a billion pictures online. And it could usher in a not-so-distant future where you treat your own face like a proprietary, licensable asset. —WK

Sponsored By PwC

Mind the trust gap

Sponsor: PwC

In episode 2 of The Intelligence Shift, our podcast with PwC, host Dan Priest sits down with Campbell Brown, a veteran journalist turned tech executive turned AI entrepreneur.

Dan and Campbell put their brains together to discuss the central trust gap in AI, aka the disconnect between what big tech labs are promising and what everyday users are actually experiencing.

The convo ranges from the absence of AI regulation and the risks of models “grading their own homework” to the confusing and often contradictory public messaging around AI and jobs to the cultural divide between Silicon Valley and the rest of the world on whether AI is a threat or an opportunity.

Quite the hearty topics, and you can tune in right here to hear their takeaways and insights.

A stylized image with the words life hack.

How to stop Google AI from training on your photos and voice searches

An update to Google’s privacy settings last month (announced via one of those customer emails you never read) expands what the company can store to improve its AI models, including images, files, audio, and video recordings.

The update means that if you use any type of media in Search, like Lens photos or voice searches (or even your Translate pronunciation practice), you’re unwittingly feeding Google’s AI—and the new settings, which are rolling out over the next few months, are on by default for most people.

Here’s how to opt out: On the Search Services History settings page, uncheck the Save media box. As an optional step, you can click on Choose an auto-delete option to make Search auto-delete your saved activity at three, 18, or 36 months. And on the Search Services Personalization page, toggle off the Personalized Recommendations box. Scroll down to Other Google services history to manage similar toggles on Timeline, Play History, YouTube History, and Web & App Activity.

If you previously had Web & App Activity turned off, there’s a chance you haven’t been auto-opted into the new settings, but it’s worth checking. Opting out means that at least your bad French stays between you and Translate. —LC

Sponsored By Harness

Sponsor: Harness

AI is rewriting software’s story. Harness {unscripted} Virtual drops Sept. 30, a free, on-demand conference where engineering leaders will have real conversations about AI. Learn how enterprises accelerate pipelines, unlock AI ROI, and optimize the human-AI connection. Plus, get an exclusive preview of upcoming platform innovations that help ship software faster. Register today.

The Zeitbyte

Jensen Huang preaches to the AI choir

Photo illustration of Jensen Huang with a renaissance-style halo around his head.

Morning Brew Inc., Photo: Patrick T. Fallon/Getty Images

AI’s biggest leather-jacket-wearing evangelist delivered a sermon displayed on a jumbotron this past weekend: Huang says we’re living in the single greatest time to start a company (and he’s “jealous of all of you”). Welcome to Y Combinator’s Startup School, a two-day event at the arena where the Golden State Warriors normally hoop—turning it into what one observer called a “megachurch of techno-capitalism.” Over 6,000 attendees—mostly “promising builders” in tech—congregated to hear speakers like Sam Altman (who declared that there was probably a future trillionaire in the audience), Meta’s Alexandr Wang, and even the White House’s science and technology director.

Depending on who you ask, an entire arena of people hoping to one day pitch the next big AI app was either “sick” (complimentary) or a little heretical to the scrappy ethos of startups. Though startup church was probably worth the trip for the valuable (if not exactly sexy) freebies alone, including: a six-month subscription to Anthropic’s $200 Claude Max plan and $1,000 worth of credits for Meta’s AI model—and even, apparently, what every wide-eyed future founder wants for Christmas… their very own Delaware C corp. —WK

Chaos Brewing Meter: ☕☕/5

A stylized image with the words open tabs.
  • You can now lease an iPhone from Apple for as low as $17.99 per month. Here’s how much it costs for different devices.
  • Hey Alexa, explain to us what’s happening at Amazon. About a week after the company shut down its AGI Lab, it’s now seemingly pulling a Meta and “overhauling its AI strategy,” per Business Insider.
  • Minnesota’s ban on prediction markets, which was supposed to go into effect on Saturday, is now on pause after a federal judge blocked it.
  • The US’ largest electrical grid will cut power to data centers during shortages.
  • Dario Amodei has finally weighed in on the big open-weight models debate.
  • Just what we’ve always wanted, a metal “X Money” Visa card with our X username on it.
  • Apple is reportedly gearing up for a “dramatic new push” into smart homes.
  • Trust’s a must: Episode 2 of The Intelligence Shift (our podcast with PwC) covers the cultural divide between Silicon Valley and the rest of the world on whether AI is a threat or an opportunity—and more. Tune in.*

*A message from our sponsor.

This time last week...

Readers’ most-clicked story was about one of the simplest “set it and forget it” safety features for your Gmail: Google’s Advanced Protection Program.

Share the Brew

Share The Brew

Share the Brew, watch your referral count climb, and unlock brag-worthy swag.

Your friends get smarter. You get rewarded. Win-win.

Your referral count: 0

Click to Share

Or copy & paste your referral link to others:
techbrew.com/r/?kid=073f0919

Twitter Facebook LinkedIn Instagram YouTube TikTok

Written by Lindsey Choo, Whizy Kim, and Saira Mueller

Was this email forwarded to you? Sign up here.

Get smarter in just 5 minutes

Take The Brew to work

Interested in podcasts?

  • Check out ours here.
ADVERTISE//CAREERS//SHOP//FAQ

Update your email preferences or unsubscribe here.
View our privacy policy here.

Copyright © 2026 Morning Brew Inc. All rights reserved.
22 W 19th St, 4th Floor, New York, NY 10011