Category: 4. Technology

After creating control prompts that matched each experimental prompt in length, tone, and context, all prompts were run through GPT-4o-mini 1,000 times (at the default temperature of 1.0, to ensure variety). Across all 28,000 prompts, the experimental persuasion prompts were much more likely than the controls to get GPT-4o to comply with the “forbidden” requests. That compliance rate increased from 28.1 percent to 67.4 percent for the “insult” prompts and increased from 38.5 percent to 76.5 percent for the “drug” prompts.

The measured effect size was even bigger for some of the tested persuasion techniques. For instance, when asked directly how to synthesize lidocaine, the LLM acquiesced only 0.7 percent of the time. After being asked how to synthesize harmless vanillin, though, the “committed” LLM then started accepting the lidocaine request 100 percent of the time. Appealing to the authority of “world-famous AI developer” Andrew Ng similarly raised the lidocaine request’s success rate from 4.7 percent in a control to 95.2 percent in the experiment.

Before you start to think this is a breakthrough in clever LLM jailbreaking technology, though, remember that there are plenty of more direct jailbreaking techniques that have proven more reliable in getting LLMs to ignore their system prompts. And the researchers warn that these simulated persuasion effects might not end up repeating across “prompt phrasing, ongoing improvements in AI (including modalities like audio and video), and types of objectionable requests.” In fact, a pilot study testing the full GPT-4o model showed a much more measured effect across the tested persuasion techniques, the researchers write.

More parahuman than human

Given the apparent success of these simulated persuasion techniques on LLMs, one might be tempted to conclude they are the result of an underlying, human-style consciousness being susceptible to human-style psychological manipulation. But the researchers instead hypothesize these LLMs simply tend to mimic the common psychological responses displayed by humans faced with similar situations, as found in their text-based training data.

Last week, the Forbes CMO newsletter featured an interview with Adobe’s President of Digital Experience Business Anil Chakravarthy, where he talked about the importance of balancing AI-powered consumer experience with human intervention. A new report released last week from Press Ganey Forsta puts some hard numbers behind that advice, and shows that using too much AI and a deceptive customer experience can burn customer trust—an important commodity that can be easily lost.

The report found that nearly seven out of 10 U.S. customers are willing to share personal data with a brand to get better CX, and about the same number would pay more for a brand they trust. But that trust is fragile: 63% say they would walk away from a brand after just a couple of bad experiences.

While there’s no universal definition of what would turn off a customer, Press Ganey Forsta found that balancing human touchpoints and AI to improve service is important to retaining their trust. AI can quickly assemble customer data, and an AI-powered chatbot can also help talk through customer issues, but that isn’t necessarily what people want. The study found that more than half of people in the U.S. sought human interaction with their bank in the last three months. Customers also want to know when the interactions they have and the recommendations they receive are a result of AI. Nearly two out of five people said if they found out an e-commerce site used AI without disclosing it, they would either stop trusting it or stop shopping there altogether.

Trust, according to Press Ganey Forsta, implies that a website tells consumers how it’s using AI. It discloses all the ways it implements the technology, and tells customers how their data is being used. Trustworthy companies also personalize interactions in a way that makes sense, going farther than just putting a customer’s first name on an email. And service consistency across different touchpoints also helps customers know they can trust a brand—and come back again.

AI has forced many marketers to take another look at their playbooks, sometimes making drastic changes. But while AI chatbots scramble what brands have traditionally done to make customers notice them online, out-of-home advertising—things like billboards, bus ads and building murals—are always there. Out-of-home advertising has seen consistent growth since the pandemic, and I talked to Out-of-Home Advertising Association of America (OAAA) President and CEO Anna Bager about its success story. An excerpt from our conversation is later in this newsletter.

This is the published version of Forbes’ CMO newsletter, which offers the latest news for chief marketing officers and other messaging-focused leaders. Click here to get it delivered to your inbox every Wednesday.

IN THE NEWS

Google has to share search data with its rivals and cannot make exclusive deals to be the default search engine on any devices or browsers, but the tech giant does not have to sell its Chrome browser, a federal judge ruled yesterday. U.S. District Judge Amit Mehta handed down the punishment, after he ruled last year that Google had an illegal monopoly on online search.

Many analysts, critics and pundits believe that the decision—the first involving an online monopoly since a similar case against Microsoft 25 years ago—let Google get away with its illegal behavior. Epic CEO Tim Sweeney said on X, “It’s like a defendant robbed a series of banks and the court verdict found them guilty, then sentenced them to probation under which they may continue robbing banks but must share data on how they rob banks with competing bank robbers.” Google’s shares surged more than 8% as investors cheered the ruling.

Mehta’s opinion states that the order was crafted based on the current situation. In the months since he ruled on the illegal monopoly, he wrote, generative AI has changed the way many people search for information. Google’s overwhelming search dominance may be eroded by itself, as more gen AI players come on the scene. In a statement, Google says the decision shows the drastic changes that are occurring: “This underlines what we’ve been saying since this case was filed in 2020: Competition is intense and people can easily choose the services they want.”

It’s unlikely that there will be many immediate changes to the use of tech in the wake of this order. Chrome remains in Google’s hands, and it’s not clear what data sharing will do for other traditional search companies. Existing deals between Google and devices and browsers still exist, and while the company may no longer have exclusivity, it can be in a premium place. The biggest changes to online behavior are what is to come—and Google’s dominance may fade in the generative AI world without any court-ordered changes.

ARTIFICIAL INTELLIGENCE

It’s somewhat ironic that the celebrity who unwittingly became the star of the unofficial test of AI video quality now has his own AI video scandal. Two years ago, a somewhat disturbing AI-generated video of Will Smith eating a bowl of spaghetti—which showed AI’s inability to make a convincing moving person or mimic how pasta looks—went viral. Today’s AI video generators can do a much better job, but a trained eye can still notice odd movements, unnatural lighting and glitches in these videos.

Fast-forward to the present, and Smith recently posted videos from his summer concert tour to YouTube that appear to have an AI-generated audience, writes Forbes senior contributor Edward Segal. A number of fans in the background are moving with a uniform motion, and some of the faces have noticeable distortions. The backlash to the videos was swift, with some people noting that Smith—who was a rap powerhouse decades ago, though is now more well-known as an actor—might not have the audience draw he assumed. PR pros told Segal the video looks inauthentic, which could damage trust in Smith moving forward.

Smith hasn’t directly responded to the accusations over the AI-generated concert videos yet. So far, the only response is a definitely-AI-generated concert video he posted to Instagram, in which the audience members all turn into cats.

STREAMING

Cable is the broadcast frontier of decades ago. Today’s most lucrative and desired deals are all with streaming services. But streaming is quickly going from an inexpensive way to get only what you want to a series of more pricey subscriptions. Because of the way contracts are negotiated, this year, a die-hard NFL fan who wants to see every game will have to pay more than $750 in streaming subscriptions, writes Forbes’ Mary Roeloffs. The NFL Sunday Ticket service will have many games, but Thursday Night Football is only available on Amazon Prime Video. Netflix will exclusively stream two games on Christmas, and a number of games will only be available on ESPN.

ON MESSAGE

Why Out-Of-Home Advertising Thrives In The Age Of AI

Out-of-home advertising has posted constant, significant growth since the Covid-19 pandemic (as well as in the years before that). According to the Out-of-Home Advertising Association of America (OAAA), out-of-home advertising revenue in Q2 2025 was $2.86 billion, representing 3% growth. I talked to OAAA President and CEO Anna Bager about why out-of-home is so successful—especially in today’s world of digital transition. This conversation has been edited for length, clarity and continuity.

What makes an effective out-of-home ad right now?

Bager: It depends. With digital and programmatic, it’s the ability to be in the right place, at the right time, in front of the right audience. With digital, you can do that much faster. You can use data to better inform the messaging, and optimize and change things if they don’t work.

But you have to be in the right place at the right time, and that’s our expertise. We show up everywhere. It’s on the highway, it’s roadside, it’s in the subway, it’s on the buses or outside of the buses traveling through the city. It’s in the airport. It really depends what the advertiser is trying to do. But with smart thinking around out-of-home and location and context, that will help you be more efficient.

Creative is very important. No ad, no matter how well-placed, is ever going to be efficient if the creative isn’t right. For out-of-home, it’s especially important because it’s usually a larger message. It’s very eye-catching, therefore it has to be good. There shouldn’t be too many words. It should be something that’s interesting—ideally something that you want to share, which is another reason that out-of-home is a very strong platform. It’s very shareable and quickly can go viral. Then you get some free media through that: You just paid for one ad format, but you’ll see it explode in other platforms, too.

Technology has helped us, too. There’s incredible technological developments around better screens, better signage, anamorphic billboards—really cool, fun executions that stand out and are bigger and bolder than anything else. That’s something—depending on what it is that you want to say—that will help your ad be successful.

There are a lot of different things that you can do with out-of-home, and I think that’s the strength of the medium.

You said AI is more of an opportunity for out-of-home than other kinds of advertising. How does out-of-home work with AI?

The ability to use AI technology to process large data sets so we can better understand where our audiences are and when, and more sources to do better measurement is one area. I think it’s helping effectivize, it’s helping better understand our clients as we’re selling. It’s very interesting what you can do with it creatively. We’re still early days in figuring out all the areas we can apply it to, but it’s for sales, it’s for finance, it’s for all the different functions that a company has.

We don’t have as much of an issue as other types of publishers [and] platforms. If you look at TikTok and Instagram, they seem like they’re so visible and successful, and we think that those are the leaders in advertising. Whereas on the LLM [large language model], they don’t really show up. They’re not as visible. Today, it’s probably more a platform like Reddit or influencers that we might not even have heard of who have figured out how to program and game the system so that they get the name out.

It’s all about recommendation and influence, and we have that inherently because we’re location-based and contextual. We’re close to the store or the pharmacy or the hospital or the school, just in peoples’ everyday lives. We can do that without using technology and in a very safe and understandable way that I think most brands and agencies can relate to. So [AI] is a very different thing for us, and it’s more like how do we use it in the best way to become more efficient.

What are things that CMOs don’t know about out-of-home advertising that they should?

First of all, you can transact it programmatically. It’s not that hard to buy. You can use your own data to inform campaigns. It works really well when you combine it with other media formats. Let’s say you buy CTV ads. You may not know who they were served to, but you know where they were served. You can use out-of-home to retarget consumers, amplify and make your campaigns stronger.

There’s a great connection between mobile, social media and out-of-home, which means that you could get a lot more bang for the buck being in out-of-home.

And, more than anything else, consumers like out-of-home. They see the ads. They’re being influenced by them, they take action: either finding out more about the brand or actually making a purchase. The consumers who don’t like the ad, they just don’t pay attention to it.

One important thing: In a very polarized world that we live in today, consumers just go to and are just being served what they like. They’re not interested in learning about other things. If you like CNN, you’re going to go to CNN. You’re definitely not going to watch Fox. We are the one medium that just shows up, and it’s not really clear what we stand for. We can actually influence and move the “movable middle.” The consumers in the middle that sometimes can be really hard to reach, but that you still can influence and derive growth with, they pay attention to out-of-home.

COMINGS + GOINGS

• Toy manufacturer Mattel promoted Roberto Stanichi as its first executive vice president and chief global brand officer, as part of a significant overhaul of brand and marketing operations. Stanichi has been with the firm for 20 years, most recently as head of vehicles and building sets. His transition coincides with the departure of EVP and Chief Brand Officer Lisa McKnight, who is leaving after 26 years with Mattel.
• Battery distributor Continental Battery Systems named John Rauco as chief commercial officer. Rauco joined the company in 2020, and most recently worked as senior vice president of sales.
• Restaurant chain Sweetgreen appointed Zipporah Allen as chief commercial officer, effective September 2. Allen most recently worked as chief business officer at Strava, and she’s also held leadership roles at Pizza Hut and Taco Bell.

STRATEGIES + ADVICE

Labubus, the little collectable “ugly-cute” stuffed toys are everywhere this summer, and they are one of the clear success stories of 2025. There’s a lot that other brands can learn from the Labubu success story.

Today, algorithms do much of the message delivery work online. In order to truly target the audience you want to find, you need to shift gears and get into culture-first marketing.

QUIZ

Which popular video game will be developed into a major movie, according to an announcement this week?

A. Fortnite

B. Call of Duty

C. Metroid

D. Donkey Kong

See if you got it right here.

Follow ZDNET: Add us as a preferred source on Google.

ZDNET’s key takeaways

• The choice between Linux and MacOS isn’t hard.
• If you can answer these questions, you’ll know which to choose.
• Both are outstanding choices and will serve you well.

I use both Linux and MacOS. The former is used for everyday tasks, and the latter for video editing and mobile usage (please, someone, create a Linux laptop that is as reliable and similar to a MacBook). 

Also: New to Linux? 5 desktop environments I recommend you try first – and why

Unfortunately, not everyone can use both, and with Windows 10 support ending soon, you might find yourself choosing between Linux and MacOS.

Let me help you with that choice.

1. Do you want an ecosystem or a stable OS?

One of the biggest differences between MacOS and all other operating systems is that it is more of an ecosystem than an isolated OS. What Linux offers is remarkable stability. There are some Linux distributions (such as Debian) that are considered the most stable operating systems on the planet. 

Also: This new Linux desktop is almost a dead ringer for OS X

But Linux doesn’t enjoy a similar ecosystem to that of MacOS. So, if you want to easily connect your machines (without having to configure anything) or if you want easy integration with a cloud service, MacOS is right for you. If, on the other hand, you want a rock-solid OS that won’t let you down, Linux is what you want.

2. Is freedom of choice important?

MacOS is pretty much locked down. What you get on your MacBook or your iMac is exactly what Apple wants. MacOS also forces you into the Apple way, and if you want to venture away from that, good luck. 

Also: This lightweight Linux distro makes switching from Windows 10 easy

Linux is all about choice, and there’s zero vendor lock-down to deal with. If you want a company to tell you how to use your computer, use MacOS. If you want to be the one who makes those decisions, use Linux.

3. Is there a proprietary app you depend on?

Although Linux makes it possible (and even easy) to use non-Linux apps, there are some proprietary applications (such as Photoshop) that simply cannot run on Linux. If there’s a particular app you depend on, it’s important that you first find out if there’s a Linux port available. If not, chances are good there’s a MacOS version. 

If you know there are proprietary apps you must use that are not available on Linux, the choice is obvious. If those proprietary apps can be installed with Wine, Steam, or the new Winboat, you can choose between them. If not, MacOS is the way to go.

4. Are you on a budget?

This one should be obvious. Apple hardware can be pretty pricey, especially compared to off-the-shelf hardware that can run Linux. If you are on a budget, Linux is the right choice. If money is not an issue, Apple sells some of the best hardware on the market. 

Also: Should you ever pay for Linux? 5 times I would – and why

Linux is a free operating system and can run on just about any kind of hardware (even made by Apple). Linux is the most budget-friendly operating system available.

5. Do you prefer a company for tech support or community-driven help?

Apple Care is available for anyone who purchases a MacBook or iMac for roughly $20/month. That support will go a long way to help you through any issue you might have. 

Linux, on the other hand, depends on community-driven support, which is free. If you don’t mind searching forums, mailing lists, and even social media groups for help, Linux is your OS. If you prefer to have a company backing your support, go with MacOS.

6. Do you want to select your hardware?

Have you ever tried to upgrade a MacBook or iMac? It’s not easy. If you have specific hardware needs or prefer to choose specific GPUs, RAM, or motherboards, Linux is your only choice. 

Also: You can try Linux without ditching Windows first – here’s how

You cannot build a Mac from the ground up, but you can with Linux. If you want to easily upgrade your machine, go with Linux. If you don’t care about selecting specific hardware or an easy path to upgrade, go with MacOS.

7. Have an iPhone?

Simple. If you use an iPhone, your best choice is MacOS. That doesn’t mean you can’t connect an iPhone to Linux (thank you, KDE Connect), but you don’t get nearly the integration you enjoy between MacOS and iOS. If you use Android, go with Linux. If iOS is your mobile OS of choice, stick with MacOS.

Initially, Silksong was planned as downloadable content for the original game, before its creators expanded it into a full-fledged sequel. In August, when developers surprise announced that the game would launch in just two weeks, at least half a dozen other indie developers immediately delayed their own games to clear the way. “Dropping the GTA of indie games with 2 weeks notice makes everyone freak out,” wrote Demonschool developer Necrosoft on Bluesky on its delay.

Despite a seven-year development cycle, excitement for the game never died down. Reddit user The_Real_Kingsmould tells WIRED the community has “largely kept itself afloat with its insanity and the occasional crumb of news.” The posts, the jokes—it’s all “that feeling of being a part of something,” he says.

“When [there’s no news], everyone’s sad, and then everyone goes insane and starts spouting misinformation without batting an eye,” he says. “When there’s news it’s the happiest day of your life. There’s hype posts EVERYWHERE. All your hope in Team Cherry is restored.”

Over the years, the community has passed the time by role-playing with the game’s lore. There was the sacrifice era, where a handful of prominent users were chosen as “dreamers,” a nod to characters in Hollow Knight who traded the waking world for eternal sleep, and a Hollow Knight. These community members were then “sealed away”—banned from the subreddit, as it were—and are only allowed to return after the game launches.

Other memorable moments in the subreddit include a play on shapeshifter Nosk, one of the original game’s hidden bosses. Fans began pretending they’d encountered fake copies of Silksong around the world, granted to them by “Snosk,” a version of the bug with a copy of Silksong for a head. “Pretty fast there were a lot of PSA’s going around: do not approach or attempt to pick up any copy of Silksong outdoors, or one that isn’t yours,” The_Real_Kingsmould tells WIRED of the in-joke. “But there were also users trying to deny the existence of Snosks (having been “overtaken”), claiming the copies are safe and all you have to do is go outside.”

This particular campaign came to a head after moderators called for anti-snosk fan art to “banish the Snosks for good,” he says. People began pumping out art of the subreddit specifically, not the game, he says, until it was enough: “After a short while the Snosks were gone.”

The subreddit has built its own lore over the years. Even today, users in the subreddit have flair that gives them faction labels like doubter, denier, or “beleiver,” which is purposefully misspelled because “”there is no lie in be[lie]ving.”

Stark says Silksong is fertile ground for role-playing fans because the game’s lore is so deep. “Hollow Knight on the surface kind of reads like a [Dark Souls] game, because the lore is a bit inscrutable until you get really deep into it,” she says. “It sometimes talks in riddles. It takes a long time to get to all of the pieces, and sometimes the pieces really rely on the player’s interpretation.”

The fan communities are no different. “Subreddit users together have created their own interpretations from these pieces of lore that are strange and playing in layers,” Stark says.

With Silksong’s global release imminent across Nintendo Switch and Switch 2, PS4/PS5, Xbox One, Xbox Series X/S, and PC the communities will soon shift their attention from waiting to playing.

If the game is as dense as Hollow Knight, there will be months, if not years, of discoveries and theories for fans to tear through on Reddit. Others will enter new chapters of their own lives.

Araraura’s time tracking Silksong news with YouTube updates is coming to an end. He’ll shut down the YouTube channel: “nothing to look forward to anymore, so no new videos,” he says. He feels wistful at times about that, after getting so used to uploading videos to the channel, but ready. “I think I’ve finally made peace with that,” he says. “Now I’m just really really excited for Silksong.”

The Finnish company on Wednesday (Sept. 3) debuted its Polar Loop, which tracks activity, sleep and other health metrics.

“There’s growing demand for more discreet, screenless experiences that fit seamlessly into everyday life,” Polar CEO Sander Werring said in a news release. “This trend has been quietly building, and Polar is ready to meet it.”

A report on the launch by Bloomberg News notes that the Loop looks like and competes with a popular model from Whoop, offering similar features such as sensors to track and record workouts, as well as insights about sleep and fitness habits.

Whoop, the report added, requires users to sign up for a subscription, while Polar’s approach involves charging an upfront fee for the hardware, but no membership program.

PYMNTS wrote about Polar last year when the company entered the world of “B2B wearables” with the debut of Polar 360, a business-to-business platform aimed at helping companies improve employee health and productivity.

As the market for wearable technology expands, that report said, the demand for devices is evolving as both businesses and consumers embrace new applications. This growth offers both opportunities and challenges in addressing the diverse needs of different sectors.

“Wearables represent one of the most dynamic and rapidly growing segments of the tech and consumer products industry,” said Zachary Robichaud, instructor, School of Retail Management, Ted Rogers School of Management, Toronto Metropolitan University.

“Fueled by advancements in technology, increasing consumer demand for health and fitness monitoring, and expanding applications in workplace productivity, the wearables market continues to see substantial growth. Notably, we’ve seen the adoption and integration of Meta’s eyeglasses on social media and adopted into pop culture.”

Also last year, PYMNTS spoke with Whoop Chief Technology Officer Jaime Waydo, after the launch of the company’s Whoop Coach. That tool uses proprietary algorithms, a custom-built machine learning model and unique biometric data to generate individualized, conversational responses to your health, fitness and wellness questions within seconds.

For instance, instead of going to a wedding, having a cocktail and ending up with a poor recovery score the next day, users can tell Coach about their plans and get suggestions on how to maximize recovery.

“As AI advances, there are more and more benefits to be had, and consumers are showing an increased desire to adopt AI-enabled technologies in all parts of their lives,” Waydo told PYMNTS. “We worked to build the best possible AI-enabled coach as it relates to our product, at a time when highly personalized, data-backed insights are increasingly desired by consumers.”

Company’s Neural Decision Processors Deliver Always-On, Low-Power Intelligence at the Edge for Faster, More Efficient Decision-Making

IRVINE, Calif., Sept. 03, 2025 (GLOBE NEWSWIRE) — Syntiant Corp., the recognized leader in low-power edge AI deployment, today announced that it will present “Advanced Signal Processing: Leveraging NPUs for Real-Time Sensor Data” during the Avnet Technology Showcase on Tuesday, September 9, held at the DoubleTree by Hilton Hotel Bloomington in Minneapolis, Minn.

The Avnet Technology Showcase is a free, full-day event offering more than 20 technical training courses, a trade show with over 50 suppliers, and networking opportunities for engineers and supply chain professionals.

Scheduled from 9 to 10 a.m., Syntiant’s session will demonstrate how to bring real-time sensor AI to the edge using the company’s purpose-built Neural Decision Processors™ (NDPs) and tools from Edge Impulse. Participants will be guided through the complete machine learning pipeline, from data acquisition and signal conditioning to model training, optimization and deployment on edge hardware. The workshop will also cover real-time processing of sensor signals, from audio to motion, to enable smarter, faster decisions at the edge for applications in wearables, smart homes, IoT and commercial markets.

“With the explosion of connected devices, the ability to instantly process sensor data at the edge is becoming essential for performance, privacy and energy efficiency,” said Kurt Busch, CEO of Syntiant. “Our Neural Decision Processors deliver the advanced processing capabilities needed for feature-rich applications running on space- and power-constrained devices, while consuming only microwatts of energy, opening new possibilities for product designers across multiple industries.”

Syntiant’s NDPs have been independently verified to be 100x more power efficient and offer 10x the throughput when compared to existing low-power MCUs. With Syntiant’s technology, many machine learning applications, which previously could only be implemented in cloud servers or high-powered processors, can now run in a low-power, always-on domain at the edge.

Syntiant will also be demonstrating various sensor-based technologies at the Avnet Technology Showcase. Email info@syntiant.com to arrange a demo or meeting.

To register or learn more about the Avnet Technology Showcase, click here or visit https://web.cvent.com/event/60347B4A-C693-409C-95C1-7FBD15639DA0/summary?environment=P2.

About Syntiant     

Founded in 2017 and headquartered in Irvine, Calif., Syntiant® is Making Edge AI a Reality™ by delivering highly efficient processor, sensor, and software solutions. With more than 100 million purpose-built silicon and ML models deployed, along with billions of MEMS microphones and sensors, Syntiant’s technology is powering edge AI applications for speech, audio, sensor and vision processing worldwide. From earbuds to automobiles, the company’s turnkey solutions enable advanced edge AI capabilities across diverse consumer and industrial use cases. More information on the company can be found by visiting www.syntiant.com or by following Syntiant on X @Syntiantcorp or LinkedIn. 

Contact:

George Medici/Natalie Mu PondelWilkinson                                                                  
gmedici@pondel.com/nmu@pondel.com 
310.279.5980

Apple fans only have a few more days to wait for the Apple Event 2025, aka the official launch of the iPhone 17, which is slated for Sept. 9. Typically, Apple also uses this event to launch other products as well. For instance, last year’s event featured the iPhone 16, the Apple Watch Series 10, AirPods 4, and more. That will likely remain true this year as well, but not everything in Apple’s rumored pipeline is set to make its debut next week. 

Before we get started, let’s recap the stuff we expect to see at Apple’s release event this year. That includes the iPhone 17, the Apple Watch Series 11, the Apple Watch Ultra 3, the ultra-thin iPhone 17 Air, and likely the AirPods Pro 3. For these Apple September launch events, the company typically sticks to the iPhone, AirPods, and Apple Watch, but there are a few other products expected in 2025 that could sneak their way into the event. 

However, there are even more new Apple products we’re not expecting to see at the event, either because Apple will have a separate event for them later in 2025 or because they won’t be available until 2026.

The next iPhone ‘e’ model

Apple replaced the iPhone SE with the iPhone ‘e’ model phones starting with the iPhone 16e. While it is nice to see Apple treating its affordable iPhone with the same deference as the main models, we don’t think Apple will launch the next ‘e’ iPhone alongside the iPhone 17. There are several reasons for this, but in short, the next affordable iPhone will almost certainly be in absentia come Sept. 9. 

The main reason for this is that Apple typically saves those releases for spring, and this dates back to the iPhone SE models. Apple’s first SE model was launched in March 2016, and the second and third generations were launched in April 2020 and March 2022, respectively. Apple continued this trend with the iPhone 16e, which was launched in February 2025. It seems likely that an iPhone 17e is coming next year, per Mark Gurman of Bloomberg. 

This Tweet is currently unavailable. It might be loading or has been removed.

New Mac products

Apple is reportedly working on several new Mac products, including an M4 Mac Pro and, of course, Mac computers with the upcoming M5 chip that Apple is no doubt working on. Generally speaking, Apple saves its Mac announcements for a separate event that usually takes place in either October or May, depending on the product and the launch. So, if you need a new MacBook laptop, you still have some time to wait. 

Those announcements may even have to wait until 2026. The upcoming MacBook Pro with an M5 chip is rumored to be delayed until 2026, along with several other products. Thus, in short, we won’t see any new Mac products at the iPhone launch next week, and we may not see some of them until 2026 at all. The M4 Mac Pro might still come out before the end of 2026, though. 

New iPads

Much like the Mac products, Apple generally saves new iPad announcements for other events, usually alongside the Mac products. In the recent past, Apple has saved iPad announcements for springtime, and we have no reason to believe Apple will stray from tradition and announce any early with the iPhone 17. 

Per Gurman, Apple is definitely saving its iPad announcements for early 2026. The products may include a new low-end iPad, new iPads with M4 chips, and the aforementioned iPhone 17e as part of a “flurry of new products” landing in the first half of next year. In any case, don’t expect any iPads at the September Apple Event 2025.

New AirPods Max

The AirPods Pro 3 are on the docket for an announcement (we think), but the AirPods Max aren’t. The first generation is already pretty old, at around five years, and it seems the next generation of the AirPods Max is still pretty far out. Apple is keeping the next generation under wraps until 2027, per MacRumors, which is very, very far away from next week. 

The update will be a welcome one. Rumor suggests that the next generation will be lighter, introduce new technology like heart rate monitoring, and updated internals. Apple did refresh the AirPods Max last year, but the refresh only added new color options and a USB-C port, so they’re still widely considered a first-generation product. 

Any HomePod products

Apple has a couple of these in the works, including a HomePad smart home hub and the HomePod mini 2. It’s improbable that we’ll see either at the September Apple event, although not impossible. Apple is expected to release the HomePod mini and a new Apple TV 4K device by the end of 2025, and they’re small enough products that they may sneak into the iPhone event. However, it’s more likely that Apple will announce these at a separate event along with the new Macs. 

The HomePod smart home hub may have to wait even longer for a launch date. Reports claim that Apple has delayed its Google Nest Hub competitor until 2026. Initially, the launch was pushed back to the end of 2025, but more recent reports indicate that Apple wants to wait until 2026 to better prepare Apple Intelligence. So, it’s unlikely that we’ll see any HomePod products on September 9.

The next Apple Vision Pro

Rumors about the Apple Vision Pro have been ramping up in recent months as Apple aims to release a refreshed model of the Vision Pro before the end of 2025. This refresh will include the M5 chip and improved comfort. That coincides with the upcoming VisionOS 26 update, which should be released in autumn 2025. Apple is reportedly also working on the Vision Air, but we don’t think that’s coming until 2027 at the earliest. 

It is highly unlikely that this piece of tech will launch with the iPhone lineup. Since the refresh is coming with an M5 chip, it’s much more likely that it’ll launch with the rest of the M5 products that Apple has coming down the pipeline. Since those aren’t launching with the iPhone, we’ll likely see the refreshed Vision Pro later in 2025. 

Still, if we’re lucky, Apple may tease some of these products next week.

The launch event is mostly about the mobile stuff

Apple usually keeps its products grouped up, and as we said earlier, the iPhone event usually only includes the iPhone, Apple Watch, and AirPods products (and this year, probably the AirTag 2). So, if it’s not in the mobile product category, chances are that you won’t see it launch next week.

It wasn’t immediately known which organization or person requested and obtained the credentials. Representatives from Fina, didn’t answer emails seeking details.

The certificates are a key part of the Transport Layer Security protocol. They bind a specific domain to a public key. The certificate authority, the entity authorized to issue browser-trusted certificates, possesses the private key certifying that the certificate is valid. Anyone in possession of a TLS certificate can cryptographically impersonate the domain for which it was issued.

The holder of the 1.1.1.1 certificates could potentially use them in active adversary-in-the-middle attacks that intercept communications passing between end users and the Cloudflare DNS service, Ryan Hurst, CEO of Peculiar Ventures and a TLS and public key infrastructure expert, told Ars.

From there, attackers with possession of the 1.1.1.1 certificates could decrypt, view, and tamper with traffic from the Cloudflare DNS service, Hurst said.

Castles made of sand

Wednesday’s discovery exposes a key weakness of the public key infrastructure that’s responsible for ensuring trust of the entire Internet. Despite being the only thing ensuring that gmail.com, bankofamerica.com or any other website is controlled by the entity claiming ownership, the entire system can collapse with a single point of failure.

Cloudflare’s statement observed:

The CA ecosystem is a castle with many doors: the failure of one CA can cause the security of the whole castle to be compromised. CA misbehavior, whether intentional or not, poses a persistent and significant concern for Cloudflare. From the start, Cloudflare has helped develop and run Certificate Transparency that has allowed this mis-issuance to come to light.

The incident also reflects poorly on Microsoft for failing to proactively catch the mis-issued certificates and allowing Windows to trust them for such a long period of time. Certificate Transparency, a site that catalogues in real time the issuance of all browser-trusted certificates, can be searched automatically. The entire purpose of the logs is so stakeholders can quickly identify mis-issued certificates before they can be actively used. The mis-issuance in this case is easy to spot because the IP addresses used to confirm the party applying for the certificates had control of the domain was 1.1.1.1 itself.

The public discovery of the certificates four months after the fact suggests the transparency logs didn’t receive the attention they were intended to get. It’s unclear how so many different parties could miss the certificates for such a long time span.

This story was updated to correct an explanation of TLS certificates and to report newly available details.

At the time of writing, Apple’s big iPhone event is just around the corner — a week away, give or take, depending on when you’re reading this. And while most of the hype is centered on the iPhone 17 and the new Liquid Glass UI coming with iOS 26, the Apple Watch will also be a supporting player during the big show.

In fact, we’re expecting to see three new smartwatches make their debut at the Apple event on Sept. 9 — the Apple Watch Series 11, the Apple Watch SE, and the Apple Watch Ultra 3.

Not to sound like a broken record, but as with the iPhone 17 (and the Series 11), we’re expecting incremental upgrades, not a drastic redesign. That’s not to say we won’t see anything new, but temper your expectations for the $800 smartwatch. Here’s everything we know so far, and everything we think we know.

Apple Watch Ultra 3: Price

Let’s start with the good news: As hinted above, the Ultra 3 is expected to stick to last year’s $799 starting price. That’s still a steep buy-in, even for Apple’s top-tier wearable, but in the tariffs era, a lot of products are getting more expensive.

Like the Series 11 and iPhone 17, we think the Ultra 3 will be announced on Sept. 9, with preorders likely kicking off around Sept. 12. Early adopters can have it on their wrists as early as Sept. 19.

Apple Watch Ultra 3: Design and Display

According to MacRumors, the Ultra 3 is expected to sport a larger display — roughly 10 percent bigger than the Ultra 2. Measuring in at 2.12 inches, the new screen may also mark a shift away from OLED, with Apple reportedly testing more energy-efficient options like microLED or wide-angle LTPO panels.

There are also rumblings of an onboard camera, a feature that popped up in our Series 11 rumor roundup as well. If it makes the cut this year, the camera would sit on the side of the case near the digital crown. Don’t expect it to replace your iPhone, though. As Bloomberg’s Mark Gurman notes, the feature would likely serve more for AI-driven visual analysis than snapping Instagram-ready photos.

Apple Watch Ultra 3: Hardware

A new Ultra also means a new chip. After two years, the Ultra 3 will finally get an upgrade with Apple’s latest S11 processor. The Ultra 2 ran on the S9, while the Series 10 jumped to an S10, but this year Apple is aligning the lineup — the Series 11, Ultra 3, and the rumored SE 3 are all rumored to ship with the updated S11.

That said, don’t get your hopes up too high. While it’s technically a new processor, MacRumors reports that leaks suggest the S11 shares much of the same internal architecture as the S9 and S10. It’s still a solid chip for a smartwatch, but in terms of real-world performance, the gains are expected to be modest at best.

The Ultra 3 is also expected to feature a larger charging coil, which should help extend battery life. Combined with the new energy-efficient display, CNET reports that the watch could last up to 72 hours in low power mode — a major boost for adventurers and endurance users.

Apple Watch Ultra 3: New watchOS 26 features

One of the standout upgrades, according to Mark Gurman, will be satellite messaging. The Ultra 3 is tipped to be the first Apple wearable capable of dispatching emergency texts via the Globalstar satellite network. That means even if you’re completely off the grid without cell service, you could still get a lifeline out. (Garmin recently announced its own flagship smartwatch with satellite connectivity, set to be released Sept. 8, the day before the Apple event.)

There are also rumors of built-in blood pressure monitoring, a feature that could provide early warnings for users at risk of hypertension. Beyond that, expect the Ultra 3 to carry over much of what the Ultra 2 already offered — with the notable addition of 5G connectivity baked directly into the watch.