It’s a struggle so many parents face today: Should I let my kid have a screen? If so, when? That decision inevitably opens a Pandora’s box of potential problems, and this relatable plight is at the heart of Toy Story 5, as pull-string cowgirl Jessie’s current child, sweet little Bonnie, is growing up in a tech-obsessed world that expects her to keep up. Will Bonnie put her childish ways behind her and lose part of herself in the process? And will Jessie lose another owner and wind up in a donation bin once again?

The answers are quite emotional, but often quite funny and extremely smart. This is Andrew Stanton’s first time back in the Pixar director’s chair since Finding Dory ten years ago, and it has been worth the wait. He takes us on an irresistible journey that honors these beloved characters, as well as the unique world where playthings live their different identities, interact with one another, and put the needs of their children above all else.

The visuals here are truly jaw dropping. It doesn’t make logical sense, but these computer generated images somehow seem more real than live action. The bumpy drips on a painted wall, the texture of cloth, and the differently smooth surfaces of plastic all look and presumably feel exactly as they should. Levels of focus within the 1.85:1 frame further sell the illusion that these walking, talking toys were actually filmed on a real set somewhere.

Colors are vibrant, and highlights are outstanding, as when the new and improved space ranger Buzz Lightyear glows in the night.

The sonic embrace of a team of heroic Buzzes is always reassuring, his deep voice coming at us from all around as he takes charge of his latest “mission.” The soundfield comes alive with ample atmospheric effects, while Randy Newman’s score enjoys a pleasing mix across the speakers. Without giving away too much, events in Act III provide ideal and thrilling fodder for the height channels in particular. I’m glad I make a habit of confirming the audio options before the movie starts, however, because this disc defaulted to Dolby Digital Plus (remember that?) rather than the supplied Dolby Atmos track.

We can tune into a commentary with co-director and co-writer Kenna Harris and others on the 4K disc, while the HD Blu-ray adds five deleted scenes, a featurette about the toys and characters cut from the final film, a gag reel, and more. Not included on either disc, but available digitally, are the Taylor Swift music video and lyric video for her song in the movie, “I Knew It, I Knew You.” A Movies Anywhere code is included.

The 4K edition is available exclusively as a SteelBook with very attractive cover artwork and a wonderfully subtle image on the reverse.

Toy Story is back. After the disappointment of Toy Story 4, this fifth (!) feature film in the franchise defies the odds and delivers everything we’d want: heart, humor, new and established characters, and a fresh, compelling story.

The movie and the disc come highly recommended.

Movie Details:

• STUDIO: Disney/Sony
• FORMAT: Ultra HD 4K Blu-ray (September 22, 2026)
• THEATRICAL RELEASE YEAR: 2026
• ASPECT RATIO: 1.85:1
• HDR FORMATS: Dolby Vision, HDR10
• AUDIO FORMAT: Dolby Atmos with TrueHD 7.1 core
• LENGTH: 102 mins.
• MPAA RATING: PG
• DIRECTORS: Andrew Stanton
• STARRING: Tim Allen, Joan Cusack, Tom Hanks, Conan O’Brien, Greta Lee, Scarlett Spears

Our Ratings

★★★★★★★★★★ Movie

★★★★★★★★★★ Picture

★★★★★★★★★★ Sound

★★★★★★★★★★ Extras

Where to buy:

OpenEvidence, an AI tool used by American doctors, is now being offered in countries with limited access to medical literature.

Reuters reported that the company is working with Anthropic to give clinicians in low- and middle-income countries free access to a specialized version of its platform.

The initiative is being introduced in around 100 countries, including Uganda, Angola, Sudan, Haiti, and Mongolia, according to OpenEvidence, which provided the information to Reuters.

The financial terms were not disclosed by either company.

OpenEvidence answers doctors’ clinical questions using peer-reviewed research and treatment guidelines. It is currently free for clinicians in the United States and Europe.

The companies say Anthropic provides the backend technology, while OpenEvidence adapts the system for each region’s healthcare needs.

Founder Daniel Nadler told Reuters that US clinicians used the platform 42 million times in August.

This year, OpenEvidence stated that it has been working with health organizations in Rwanda and Botswana to modify its tools for situations in which disease patterns.

The resources available for diagnosis and the treatments on offer are very different from those found in richer countries; the project in Rwanda, carried out in partnership with the Rwanda Biomedical Center and Resolve to Save Lives, involved 45 local doctors and nurses who tested the tool and provided feedback on any necessary changes.

“One hundred percent of what we are developing for these areas is designed to be context adaptive,” Nadler told Reuters.

Critics of medical AI warn that systems trained mainly on data from high-income countries may not reflect how medicine is practiced elsewhere.

In response, Nadler points out that even in places where clinics do not always have electricity, most doctors still have smartphones.

Daniela Amodei, president of Anthropic, said the project is something the market could not achieve on its own.

In her interview with Reuters, she explained that while the technology is now advanced enough, “market incentives by themselves would not cause it to happen” without some philanthropic effort.

Also, we reported this week that Anthropic has built a laboratory for physical biology as it expands into drug science.

For clinicians using the tool, the main issue is access, not AI itself. Dr Ahmed Bendary, a cardiologist at Benha University in Egypt, told Reuters that expanding the tool beyond the United States and Europe is especially important in places where access to major journals is limited.

The key question now is whether a tool built mostly on evidence from wealthy countries can be adapted for use in places like Mongolia and Sudan, one country at a time.

The keynote kicking off Snapdragon Summit 2026 in Maui had the latest forecast for how AI agents will organize our lives through the phones in our pockets. Qualcomm CEO Cristiano Amon spent some time walking viewers through how the agents will buy things for them, preparing for the day when AI will have access to our bank accounts or credit cards — and why that’s a good idea.

For years, Qualcomm has pitched a future where AI agents on your devices know your habits and preferences well enough to make informed decisions on your behalf. Eventually, this will include the capability to spend your money on things such as necessities, gifts and travel. The question is how agents will decide whether or not to pull the trigger on a purchase — and how much input you’ll have on what decisions they make.

Nobody’s figured this out yet, and might not for some time. The fear is that an AI agent would run off with your bank account numbers, clear out your savings and run up your credit card. But at this year’s summit, Qualcomm suggested how this might actually work, wherein agents would repeatedly check in before they spend your money.

“When an agent is acting as a proxy of you, you need to trust the agent, and you need to trust the device, especially for something like shopping,” Amon said on the keynote stage. 

Naturally, the CEO was pitching a method with his company’s technology at the center, but it was at least sensitive to the uncertainties people likely have about an AI buying things for them. He said that the company is developing technologies to protect personal information and provide a secure device environment.

“The agent needs to authenticate and operate and keep you in control,” Amon said. 

Around 45 minutes into the keynote, Qualcomm presented a way this might go. A woman wearing smart glasses walks past a store when the AI agent suggests that, in the past, she’d bought sunscreen and lip balm from that location — did she want to buy more? She okays the purchase, and the AI agent says she can go pick up her goods. She then proceeds down the street and asks the agent about getting a laptop, and the AI agent guides her to a shop where she can view several on display. She decides on one, but the agent asks her to confirm the purchase first.

I didn’t love everything about the demo, as her purchase was followed by a reminder about her credit card rewards points in what felt like a nagging ad, but that’s probably a personal preference I could tell my AI agent to avoid. From a consumer standpoint, the most important part was waiting for confirmation before making a purchase — especially when plunking down several thousand dollars for a pricey computer. 

Qualcomm is treading carefully here, and shrewdly brought in execs from a trusted purchasing platform to reassure viewers that the financial institutions they already trust will be part of the process. MasterCard CEO Michael Miebach, speaking in a prerecorded video, reaffirmed that people will only use processes they trust to keep their payments and data secure; on stage, Amon was joined by MasterCard senior fellow Raja Rajamannar, who spoke plainly about the implicit trust that an AI agent would work on their behalf — which will come from a brand they already trust, such as Mastercard. 

This is all self-promotional, but they’re speaking to relatable uncertainty. Over the past decades, people have become a lot more comfortable with online shopping, with the initial skepticism thawing as the process became trustworthy. AI agents further separate people from the act of handing over their hard-earned cash, so another bridge of trust must be built. Envisioning the next process with clear moments when an AI agent will stop to ask people before making purchases on their behalf is reassuring amid uncertainty about a new way to spend money.

The long road to letting AI agents pay for you

The technology at the center of the process is the AI agent itself, which runs on your smartphone. In a media roundtable discussion after the keynote, Chris Patrick, Qualcomm SVP and general manager of handsets, emphasized that building consumer trust in the process would take time.

Dino Bekis, Qualcomm SVP of premium mobile and AI business unit, reasoned that the answer will include a lot of reassurance by making financial institutions a central part of the process. Just as institutions now take care of fraudulent charges or other mistaken purchases when your credit card is stolen, there will need to be reassurances if AI agents wrongly buy something that consumers didn’t approve.

“Financials [will need to] do some level of backstop to say, hey, you as a consumer, if this happens, there’s going to be some level of consumer protection for you, right?” Bekis said.

Later on, I wandered into a demo room where Qualcomm showed off concepts revealed on stage, one of which simulated how an AI agent would buy something for you. The demonstration followed three clear steps: identifying something to buy, the AI agent finding a relevant suggestion to ask about buying it and the actual purchase process. 

Qualcomm’s Francisco Cheng, senior director of product marketing, wearing off-the-shelf TCL Rayneo smart glasses, looked at a poster for an experience (aptly, a Hawaiian volcano). The AI agent recommended a particular tour of a nearby volcano, and when Cheng confirmed that he’d like to buy it, the agent — running on a connected smartphone — handed the transaction over to Mastercard to make the purchase. Any disputes could, in theory, be resolved by the credit card company.

“I think if there’s a level of mistrust by letting an agent handle things through and through, at least Mastercard can step in as a trusted payment provider to handle that piece of it,” Cheng said.  

People will still surely be gunshy about the whole process, Cheng admitted, but at least there are trusted brands handling every step of the process. And that’s exactly how it went with online shopping. 

“I’m old enough to remember when I was skeptical about making online payments, about doing banking transactions online. I was comfortable writing a check to somebody,” Cheng said. “But today, I wouldn’t even think twice about paying bills online. It’s a learning process.”

David Lumb

Managing Editor, Mobile

David Lumb is a managing editor for the mobile team, covering mobile and gaming spaces. Before CNET, he covered tech, gaming, and culture for TechRadar, Engadget, Popular Mechanics, NBC Asian America, Increment, Fast Company and others. As a true Californian, he lives for coffee, beaches and burritos.

See full bio

Advertisement

So close and yet so far…