Join Mikko Hyppönen and security leaders from the NFL, CHANEL, and Atlassian for a two-hour digital summit on what AI-speed attacks change, what defenders should stop doing, and how to validate, decide, fix, and re-validate at machine speed.
Tech
Why is Soup Restaurant losing money?
35-year-old Soup Restaurant has swung from highs and, most recently, losses
For more than 30 years, Soup Restaurant has been a familiar name in Singapore’s dining scene. Many Singaporeans would recognise it for its double-boiled soups, home-style Cantonese dishes and, perhaps most famously, its Samsui Ginger Chicken.
The business began in 1991 with a single Chinatown shophouse, six staff and six dining tables.
Today, the company behind it, Soup Holdings Limited, is listed on the Singapore Exchange and runs 15 outlets across Singapore and Malaysia, alongside two franchised outlets in Indonesia, as well as central-kitchen operations, sauce distribution and a social enterprise.
But the company’s latest numbers tell a very different story from its expansion years.
In the first half of 2026, Soup Holdings swung to a S$673,000 loss, from a S$224,000 profit a year earlier.
And over the last two years, the group has closed outlets and cloud kitchens, impaired assets linked to underperforming businesses, and begun rationalising parts of its portfolio.
So how did a restaurant that’s been part of Singapore’s dining landscape for more than three decades get to this state?
An opportunity spotted by four engineers
Soup Restaurant was founded as a side project by four engineers: Mok Yip Peng, Wong Wei Teck, Wong Chi Keong and Ho Hong Chin.
In 1995, the business introduced its Samsui Ginger Chicken, inspired by the Samsui women who came to Singapore as migrant workers. The dish—poached chicken served with a ginger-and-scallion dip—became one of the restaurant’s signatures.
But the company soon realised that if it wanted to grow, it couldn’t remain a small Chinatown restaurant, so it took a relatively measured approach to expansion.
Its second outlet only opened in 1997, six years after the original. Then, in 1998, it opened at Causeway Point, bringing the brand into Singapore’s growing suburban mall culture.
This was an important transition for the brand. Instead of merely relying on Chinatown’s tourist traffic and heritage appeal, Soup Restaurant could now reach families and office workers in neighbourhoods around Singapore.
The company kept opening stores until the outlet count reached 10, and eventually developed centralised food-processing capabilities, allowing some preparation to be standardised across its restaurants.
Centralisation made the business more consistent and even more scalable. By the mid-2000s, Soup Restaurant had become a sizeable group—and in 2007, Soup Restaurant Group Limited—the company that would later become Soup Holdings—listed on the Singapore Exchange.
Building an integrated F&B ecosystem
As the business grew, so did its ambitions. Soup Holdings expanded beyond its original restaurant concept and built an entire portfolio of F&B businesses.
One of the more notable additions was Dian Xiao Er, known for its herbal roast duck. The group also developed Teahouse by Soup Restaurant, serving Nanyang-inspired dim sum, and Cafe O, a halal-certified local coffee shop concept running since 2013.
It also began moving further upstream and downstream in the food business.
The group invested in food processing, distribution and procurement operations, while Soup Restaurant products such as bottled sauces and ready-to-eat meals allowed the brand to extend beyond its physical restaurants.
Instead of relying entirely on one restaurant brand, Soup Holdings could operate different concepts for different customers and occasions, reaching more of them through both business and consumer channels.
And for a while, that strategy worked.
Crossing S$40 million
Soup Holdings’ revenue grew substantially from its early years, eventually climbing into the S$40 million range in the years just before COVID-19.
The pandemic then hit restaurants particularly hard. Soup Holdings’ revenue fell to S$29.6 million in FY2021, with dining-in restrictions and safe-management measures weighing directly on sales.
The company survived the downturn, and as Singapore reopened, revenue climbed back—to S$37.6 million in FY2022, then S$41.1 million in FY2023. For a moment, it looked like the group was clawing its way back to pre-COVID levels.
Then, it reversed again. Revenue fell to S$38.4 million in FY2024 and S$37.5 million in FY2025—and, more tellingly, profitability deteriorated sharply alongside it.
Soup Holdings posted a S$1.8 million profit before tax in FY2023. A year later, that had flipped to a S$3.1 million loss. The loss then narrowed 63.3% to S$1.1 million in FY2025, but the group remained in the red for a second straight year.
That makes the current situation more complicated than simply saying Soup Restaurant has lost customers.
The group is still generating tens of millions of dollars in annual revenue. The bigger question is why that revenue is no longer translating into sustainable profits.
The costs of growing an F&B empire
Soup Holdings’ recent results reveal something easy to miss when looking only at the number of outlets.
Every new restaurant brings in revenue. But it also brings rent, salaries, renovation costs, equipment, utilities, food costs and lease commitments.
If an outlet performs well, higher sales can absorb those costs. If it doesn’t, the outlet can become a drag on the wider group—and Soup Holdings’ disclosures suggest that some parts of its expanded portfolio had reached that point.
For FY2025, the company recognised S$400,000 in non-cash impairment losses on plant, equipment and right-of-use assets tied to three underperforming business units. In practical terms, that meant the group had to write down assets associated with parts of the business that were no longer performing as expected.
That has pushed the group towards a more selective approach to expansion.
In 2025, that meant closing two outlets, including a 23-year-old outlet at Compass One, and four cloud kitchens, resulting in S$1.3 million in lost revenue. The lease expiries of two other outlets affected a further S$1.1 million in revenue.
At the same time, Soup Holdings was still opening new locations. Two new outlets—one opened at the start of 2025 and another at The Star Vista in Nov—generated S$2.6 million in revenue.
The approach, then, isn’t simply to stop expanding.
The group has described it as retaining and refocusing on outlets with sustainable turnover potential while exiting operations that no longer meet its performance benchmarks. It is also renovating and refreshing the outlets it retains, although those renovations could temporarily weigh on FY2026 performance.
But choosing which outlets to keep is only part of the challenge. The environment in which those outlets operate has changed significantly.
Operating in Singapore’s challenging F&B landscape
Soup Restaurant’s original proposition was built around a simple idea: people wanted good, familiar Chinese food they could share with family and friends.
That proposition hasn’t disappeared. But the market around it has become far more crowded, with independent restaurants, international chains, delivery-first concepts, food courts, cafes and newer F&B brands all competing for consumers’ spending.
And the cost of running a restaurant has risen too.
Manpower remains a persistent challenge, food costs fluctuate, rental costs can make or break an outlet, while consumers have become more price-sensitive as comparing restaurants and promotions online has become easier.
Soup Holdings itself has pointed to intense competition from local and foreign operators, rising operating costs, manpower constraints and changing consumer preferences, and even flagged the upcoming Johor Bahru-Singapore Rapid Transit System Link as a potential swing factor for cross-border travel and F&B spending across both markets.
For a mature restaurant group, these pressures can add up quickly. Higher costs eat into margins, while a competitive market makes it harder to simply raise prices to compensate.
And Soup Holdings isn’t relying solely on its restaurants.
Its food processing, distribution and procurement segment was meant to provide another source of revenue beyond dine-in customers. But that business came under pressure too: segment revenue fell by S$600,000 in FY2025, which the group attributed to expired contracts for supplying ready meals, amid market saturation and challenging macroeconomic conditions.
So is Soup Restaurant in decline?
The first half of FY2026 offers perhaps the clearest picture yet of where Soup Holdings stands.
Revenue came in at S$19.5 million, down just S$100,000 from the same period a year earlier. On the surface, that seems like a relatively stable number from the year before, and close to maintaining that yearly S$40 million revenue figure.
But dig deeper, and the company has actually swung from a S$224,000 profit to a S$673,000 loss.
Existing outlets actually generated S$600,000 more revenue, and two new outlets contributed a further S$1.4 million—but these gains were offset by the closure of two outlets following relocation and lease expiry.
At the same time, costs rose. Purchases and other consumables climbed 0.9 percentage points to 21.1% of revenue, partly due to higher raw-material costs during the seasonal period.
Employee benefits expenses rose by S$300,000, while other expenses increased by a further S$400,000—the latter partly due to a one-off termination fee payable to a landlord after the group decided not to proceed with a proposed outlet lease.
All this illustrates the underlying problem facing Soup Holdings: when margins are already thin, relatively small increases in food, manpower and operating costs can wipe out a modest profit.
So, is Soup Restaurant in decline?
The answer isn’t quite that simple. Soup Holdings is closing underperforming outlets, cutting back on cloud kitchens, impairing assets and restructuring parts of the business. It has also recorded two consecutive years of losses, while net asset value per share has fallen to 2.44 cents from 4.25 cents in FY2023.
Yet its core restaurant business isn’t simply collapsing. Sales from existing outlets increased in H1 FY2026, while the group generated S$6.3 million in net cash from operating activities in FY2025 and continued to propose a final dividend.
The bigger issue is that the business has become harder to run profitably, while the growth opportunities that once justified expanding the group have become harder to find. For Soup Holdings, the challenge now isn’t simply growing bigger.
It’s figuring out how much of its existing portfolio the core business can sustainably support—while also finding ways to make the remaining outlets more compelling.
A brand refresh may work
Soup Holdings says it’s refreshing its brand positioning and concepts, while also restructuring its leadership and exploring AI-driven kitchen solutions to tackle manpower challenges.
That suggests the company knows cost-cutting alone won’t solve the problem.
If Soup Restaurant simply closes its weak outlets without making the remaining ones more compelling, it risks ending up as just a smaller version of the same business.
- Read other articles we’ve written on Singaporean businesses here.
Featured Image Credit: P Chan via Google Maps
Continue Reading
Just buy a basic gaming PC instead.
If you’ve been looking to get more gaming power out of your laptop but don’t want to pay for a replacement amid RAMpocalypse pricing, an eGPU could be your best bet. An eGPU is an external graphics processor, usually a desktop graphics card in an enclosure connected to a compatible laptop or gaming handheld through Thunderbolt or USB4. They can be pretty good at taking over rendering, meaning even thinner and weaker machines can run more demanding games with one plugged in, but there are many caveats.
A GPU run as an eGPU will typically operate slower over Thunderbolt than it would inside a desktop. The enclosure to house it costs hundreds of dollars too, and some devices, like Apple Silicon-based Macs, can’t use an eGPU for gaming at all. That said, eGPU hardware is in better shape today than it has been in years. Razer revived its eGPU line with a Thunderbolt 5 dock in July 2025, and Framework is building an eGPU kit for its Laptop 16.
How much performance is lost with an eGPU?
A Thunderbolt 4 port carries 40Gbps, and Intel’s spec allots PCIe data 32 Gbps of it, an eighth of the bandwidth a desktop PCIe 4.0 x16 slot provides. Most of what the card in an eGPU enclosure does will never cross the Thunderbolt cable. Textures and shaders get loaded into VRAM once and remain there, meaning the cable only gets busy when a game pulls in new assets or sends finished frames back to the display.
Portables review site Notebookcheck tested an RTX 2080 Ti enclosure back in 2021 and found deficits ranging from seven to 32 percent when compared to a desktop. Losses did shrink to single digits at 4K on an external monitor, but rendering to a laptop is slower because finished frames travel back down the same cable. In contrast, PC Gamer tested an RTX 4070 Ti and 4090 over Thunderbolt, OCuLink and desktop PCIe. Its RTX 4070 Ti ran around 25 percent slower over Thunderbolt than OCuLink in many 1440p tests, with lower 1 percent lows and more stuttering.
PC Gamer also claimed that upgrading from the 4070 Ti to the 4090 delivered smaller gains over Thunderbolt and OCuLink compared to desktop. So, the faster the card, the more of its performance the Thunderbolt cable restricts, making a mid-range GPU the more sensible choice for an eGPU enclosure. Thunderbolt 5 provides 80 Gbps of connection bandwidth, while Razer’s $350 Core X V2 connects the GPU through PCIe 4.0 x4, nominally 64 Gbps. That gives an x16 graphics card four lanes instead of the 16 available in a suitable desktop slot.
Cost and compatibility
The Core X V2 comes with only the enclosure; no power supply, USB ports or Ethernet. You’ll need to add your own ATX PSU, and Razer now sells the missing I/O separately as a Thunderbolt 5 dock starting at $390 for the Mercury version. Throw in the graphics card itself and there’s really little point in buying all these components when you could just invest in a budget gaming rig instead. At least in doing so, you’ll be avoiding the external connection’s bandwidth limit. Thunderbolt 5 laptops are also few and far between outside premium gaming machines and mobile workstations, so connecting the enclosure to a compatible Thunderbolt 4 laptop limits the connection to 40 Gbps regardless.
OCuLink does manage to get around Thunderbolt by running bare PCIe over a cable. That’s the route Framework is taking. Its OCuLink Dev Kit exposes an eight-lane PCIe connection on the Laptop 16, rated at up to 128 Gbps. You end up sacrificing some convenience here given that OCuLink doesn’t power the host, and there’s none of the plug-and-play that makes a Thunderbolt box a one-cable dock.
Apple dropped eGPU support with its move away from Intel silicon back in 2020. No macOS release since has brought it back for graphics acceleration. However, Apple did approve TinyGPU earlier this year, a driver extension from the tinygrad project that runs newer AMD and Nvidia cards over Thunderbolt or USB4. It’s for AI workloads only, and its lack of display output means it won’t get you anywhere for gaming use cases.
A new Android malware called RatHat has been discovered, targeting users with an AI-powered subsystem that helps operators remotely navigate compromised devices.
Zimperium zLabs researchers analyzed the malware and believe it is linked to threat actors from China after finding it using LLM prompts written in Chinese.
The researchers say the malware is distributed through malvertising, SMS, and phishing sites promoting APK downloads from outside Google Play.
Like most Android malware families, RatHat relies on abusing Android’s Accessibility permissions to perform highly privileged actions on infected devices.
Most notably, it enables Developer Options and Wireless Debugging to gain a local shell-level execution context without requiring an external computer, similarly to mechanisms seen in the ToxicPanda and RedHook Android malware families.
ADB access lets RatHat install a Go-based agent (liblocal-service.so) that runs commands with ADB shell privileges, creates battery-restriction bypasses, and manages persistence.
The same agent also restores the malware if it’s removed or stopped, providing a persistence channel independent of the malware itself. This back-up goes both ways, with the malware restoring the agent if it’s removed.
Source: Zimperium
A second agent (libmedia_codec.so) also acts as an FRP reverse-proxy client, establishing a persistent tunnel to the attacker.
The malware displays HTML overlays for targeted banking and cryptocurrency applications to capture account credentials.
It can also intercept SMS messages and notifications, including one-time passwords; record text-change events; extract URLs from browser address bars; and capture lock-screen PINs, passwords, and unlock patterns.
Source: Zimperium
AI-guided navigation
Zimperium reports that RatHat uses an AI-powered user interface-automation engine that serializes the live Android Accessibility tree into XML and sends it to a popular AI assistant not named in the report.
The AI tool is then tasked with:
- Identifying the center coordinates of a named interface element
- Determining an element’s actual on-screen text
- Returning navigation instructions such as ‘SCROLL_DOWN’ and others
This AI-powered subsystem makes RatHat’s interface automation more adaptable than automation based entirely on fixed scripts, allowing remote navigation without requiring the operators to perform real-time interaction.
“RatHat uses AI to intelligently navigate and control the device interface in real-time, making its operations more adaptable and harder for security software to detect than traditional, scripted automation,” comments Zimperium.
Source: Zimperium
The researchers warn that RatHat will actively thwart removal attempts by intercepting the uninstall confirmation screen, canceling the procedure, and displaying a fake Google Play overlay containing a bogus error message.
Zimperium also highlights multiple anti-analysis techniques embedded in the malware, including APK container tampering, a bloated (61MB) Android manifest file, and invalid DEX pseudo instructions, all aimed at confusing or breaking analysis tools.
Android users should never download APK files from outside Google Play unless they explicitly trust the publisher, avoid granting Accessibility permissions to apps, and regularly scan their devices with Play Protect.
If you’re buying a new iPhone 18 Pro, you’re probably looking for a new charging accessory. You might want to take advantage of the higher fast-charging ceiling (60W) or declutter your desk with a multi-device charger. To help you decide, I’ve reviewed dozens of Amazon listings to find the best iPhone 18 Pro charging accessories so you can choose what’s right for you.
Wall Adapters
Apple 40W Dynamic power adapter ($34.99)
If you asked me to point you to one power adapter for iPhone 18 Pro owners, this is where I’d start, mostly because Apple’s numbers are hard to beat: a 4.8-star average from over 1,500 ratings. At $34.99, it’s also priced like a standard third-party brick, even though it’s the official one.
Why it stands out
- It’s labeled 40W but dynamically pushes up to 60W depending on what’s plugged in, so it doesn’t leave power on the table the way a flat-rated adapter would.
- USB-C output pairs with the same cable already in most iPhone boxes, so you don’t have to track down an adapter-specific cable.
- It offers unmatched peace of mind.
Anker 45W Nano smart display charger ($29.99)
This isn’t just another 45W adapter with foldable prongs. It’s the only charger on this list with an actual screen, and it’s earned a 4.7-star average across more than 5,400 ratings for that reason.
Why it stands out
- The built-in display shows real-time wattage, temperature, and charging status at a glance, genuinely useful for anyone who’s wondered if a charger is actually working.
- Care Mode lowers charging temperature by up to 9°F versus a standard 45W charger, specifically built for safer overnight charging.
- Comes in four colors.
Anker 67W 3-port GaN charger ($39.99)
Anyone juggling a new iPhone 18 Pro alongside AirPods, an Apple Watch, or a tablet should skip single-port chargers and go straight to this one: Anker’s slightly older but still well-reviewed 3-port GaN charger.
Why it stands out
- The only wall adapter on this list that fast-charges three devices at once, up to 67W from a single port, or up to 65W split across two
- Also doubles as a genuine laptop charger in a pinch, useful for anyone traveling with a MacBook and a phone but only one outlet.
- Comes with a 24-month warranty.
Vention Dash Pro 140W 4-Port GaN charger ($59.90)
Vention just launched this at IFA 2026, and it’s the one adapter that can charge your iPhone 18 Pro and your MacBook without being dramatically slow on either. Of course, you can use the other ports to charge the other devices you might have. It also has a display that shows charging information in real time.
Why it stands out
- Reads charging status through five color-coded emoji expressions instead of plain numbers.
- 140W across three USB-C ports and one USB-A beats every other wall adapter here, enough for a laptop and phone charging together.
- Supports PD 3.1, QC, PPS, and UFCS protocols, so it works across older devices too, not just the newest ones.
MagSafe Chargers
Apple MagSafe charger ($39.99)
This is the safe, no-homework option for anyone who doesn’t want to check whether a random wireless puck meets Apple’s charging spec and still get the standard MagSafe experience.
Why it stands out
- Comes in two cable lengths, 1 meter or 2 meters, an option most third-party MagSafe pucks don’t bother offering.
- Reaches its full 25W speed only when paired with a 30W+ USB-C adapter.
- Also charges AirPods with a wireless case.
Belkin 25W 3-in-1 foldable charging station ($53.99)
This is the best value on this list in my opinion. It’s the only charging station under $60 that ships with everything needed to start charging the same day.
Why it stands out
- Includes a full 45W power adapter and 5ft cable in the box, unlike most standalone MagSafe pucks that sell only the puck.
- Folds completely flat for a bag, then unfolds with an adjustable stand angle up to 65° for StandBy Mode on a nightstand or desk.
- Worth knowing: the full 25W speed needs iOS 26 or newer.
Anker Prime 3-in-1 Qi2.2 charging station ($139.98)
If the iPhone 18 Pro is joining an Apple Watch and a pair of AirPods already in daily rotation, this is the station I’d point toward, even at nearly three times Belkin’s price.
Why it stands out
- The only station on this list with a touchscreen, a 1.65-inch display showing real-time wattage, temperature, and status for all three devices at once.
- Charges the phone, Apple Watch, and AirPods simultaneously from one outlet, decluttering your desk.
- It looks good on a desk or a nightstand.
ESR MagSafe car mount charger ($18.99)
This is the only accessory in the guide built for the car, doubling as a phone mount and charger for a dashboard or air vent.
Why it stands out
- Switches between portrait and landscape with one hand, making it easy to glance at turn-by-turn directions.
- 1,600g magnetic hold plus a non-slip silicone ring, built to survive bumpy roads rather than sitting flat on a desk.
- Needs its own 18W+ car adapter for full-speed charging, sold separately.
Wireless Power Banks
Ugreen 10,000 mAh Magnetic Power Bank ($46.99)
This wireless power bank focuses on capacity. At 10,000mAh, it holds nearly two full charges for an iPhone, more than any other wireless power bank on this list.
Why it stands out
- Genuinely the highest capacity in this roundup.
- Supports simultaneous wired and wireless charging with intelligent power distribution, so plugging in a cable while the phone sits on top wirelessly doesn’t needlessly slow either one down.
- It doesn’t feel like carrying a brick despite the capacity.
Anker 5,000 mAh nano magnetic power bank ($54.99)
This is the thinnest power bank on this list by a wide margin, and Anker built it specifically as an everyday-carry item rather than a full backup battery.
Why it stands out
- At 0.34 inches thick, it’s genuinely the slimmest option here, thin enough to sit under a phone case without adding noticeable bulk.
Baseus Picogo MagSafe power bank ($55.98)
If speed across the board matters most, this power bank wins on every metric: wireless output, wired output, and recharge speed.
Why it stands out
- The fastest wireless output of any power bank here: 25W Qi2.2, enough to push an iPhone 18 Pro to 50% in about 30 to 35 minutes.
- Also works as a genuine 45W wired charger through its USB-C port, not just a wireless-only backup battery.
- Refills its own 10,000mAh battery in about 1.5 hours over a 30W connection.
Saad Nadeem doesn’t have any professional background in home improvement. But like any self-respecting weekend warrior, his experiences as a homeowner have taught him some valuable lessons.
The biggest lesson, perhaps, was to create his own startup to solve his problems.
Nadeem is co-founder and CEO of Local Crew, a newly launched Seattle startup designed to make it easier to find laborers, assess bids, and hire a professional to perform everything from yard work to house repairs.
“Even for straightforward projects, I had to contact several pros, explain the same job repeatedly, arrange on-site visits, and sometimes wait days just to compare estimates,” Nadeem said.
Professionals are equally frustrated, he learned, because traveling to provide free estimates is time consuming, particularly for smaller jobs they may not win.
@media (max-width: 600px) {
.gw-founder-box { float:none !important; max-width:100% !important; margin:20px 0 !important; padding:16px 18px !important; }
.gw-founder-box .gw-label { font-size:12px !important; margin-bottom:8px !important; }
.gw-founder-box .gw-hero { font-size:19px !important; }
.gw-founder-box .gw-byline { font-size:15px !important; }
.gw-founder-box .gw-meta { font-size:12px !important; margin-bottom:14px !important; }
.gw-founder-box .gw-facts { padding-top:14px !important; }
.gw-founder-box .gw-facts li { margin-bottom:12px !important; }
.gw-founder-box .gw-facts li:last-child { margin-bottom:0 !important; }
.gw-founder-box .gw-value { font-size:15px !important; line-height:1.45 !important; }
}
A Pakistani immigrant and tech veteran, Nadeem has spent nearly a decade working in marketplace operations at companies including Uber, Roblox, and TikTok, focusing on customer experience, support operations, and simplifying complex processes for users.
Local Crew relies on AI to analyze a video submitted by a customer in which a potential job is spotlighted. AI turns the video into a structured job post with a project description, relevant details, and a recommended price. The customer can adjust the price before posting, and local pros can accept it or submit a counteroffer. The main differentiator is Local Crew alleviates the need for multiple calls or in-home estimate visits.
Companies such as Thumbtack, Angi, and Taskrabbit are among the closest small-job and home-improvement comparisons.
Local Crew — now live on iOs and Android — is starting in the Seattle area, with plans to expand into major Midwest cities next and eventually grow nationwide.
“The goal is to prove the model in Seattle, learn from customers and pros, and then expand thoughtfully into additional markets,” Nadeem said.
We caught up with Nadeem for this edition of GeekWire’s Startup Spotlight to learn more about his startup.
In 50 words or less, give us your startup’s elevator pitch.
Local Crew is a platform that makes it simple for customers to find and hire local professionals for any task. Whether it’s homeowners who need a pro to come fix their leaking faucet, or a student looking for some extra hands to help them move. Local Crew cuts the back and forth and exhausting visits for estimates. Simply record a video of what you need done, and Local Crew analyzes your task, creates a detailed brief, and suggests a price you can set. Local professionals bid on your task and you pick the one that suits you.
What problem are you obsessed with solving?
Hiring someone for tasks is still surprisingly inefficient. Homeowners spend days coordinating visits just to compare quotes, while pros lose hours driving to estimates that never turn into jobs. We’re building a faster, simpler way for both sides to connect.
What surprised you after talking to customers?
We expected homeowners to be our biggest challenge. Instead, we learned that local pros were just as frustrated. Driving across town to provide free estimates for small jobs often wasn’t a good use of their time, especially when many estimates never turned into work. That’s when we realized we were solving a problem for both sides of the marketplace.
How has AI changed the way you build your company?
AI has accelerated everything — from product design and engineering to marketing, customer support, and content creation. It has allowed a small founding team to move with the speed and output of a much larger company.
What’s one thing people misunderstand about your startup? People often assume we’re another lead-generation platform. We’re not. Customers record their project, set the price, and pros bid or counteroffer. The goal is to eliminate unnecessary back-and-forth and quote visits before work begins.
What’s the toughest decision you’ve made in the past year?
Delaying our public launch. We could have launched earlier, but we chose to spend more time improving the experience based on real feedback from early users. It was the harder decision, but the right one.
What’s the one piece of advice you give to other entrepreneurs?
Talk to customers far more often than you think you should. The best product decisions usually come from listening, not guessing.
We’ll know our company has made it when…
Hiring someone for a home project feels as easy and expected as ordering food or requesting a ride, and homeowners naturally think of Local Crew first.
Geothermal startup Mazama Energy raised $135 million in an oversubscribed Series B, the startup said Thursday.
The fresh funding will help the company develop horizontal wells in super-hot rocks capable of generating 15 megawatts of electricity per well.
Mazama plans to start generating electricity sometime next year and said its first site in Oregon is capable of producing 10 gigawatts of electricity. That’s a sharp upward revision from the company’s estimate last year, when investor Vinod Khosla said the site had the potential to produce 5 gigawatts.
Enhanced geothermal startups are shaping up to be the dark horse in the battle to power AI data centers and other large loads on the grid. While tech companies and data center developers have been throwing their weight behind natural gas, geothermal startups have been quickly advancing their technology.
Unlike previous geothermal projects, enhanced geothermal startups are targeting deeper and hotter rock. Mazama, for example, has drilled past 10,000 feet in 15 days on its way to about 15,000 feet deep. At those depths, the company can access 750˚ F (400˚ C) heat, which at high pressures turns water into a “supercritical” fluid that’s neither fully liquid nor fully gas — a state that lets it absorb far more energy than ordinary steam. Mazama said that will allow each well to produce up to 10 times more power than traditional geothermal technologies.
Given the depths Mazama is targeting, the potential payoff is significant. Tapping just 1% of super-hot rock worldwide could generate more than 63 terawatts of electricity, according to the University of Twente and the Clean Air Task Force.
While work at its Oregon site is still in the early stages, the startup has already secured land for a second development. Mazama aims to complete development of the first site by 2030, at which point it says it should generate 200 megawatts of electricity.
The new round was led by Centaurus Capital and Doerr Capital, with participation from ConocoPhillips and Shell Ventures. Existing investors including Khosla Ventures and Gates Frontier returned, and new investors SiteGround Capital, H. Barton Asset Management, and the Jeffrey and Marieke Rothschild Foundation also joined the round.
Mazama was incubated at Khosla Ventures.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Can You Feel It: Haptic feedback has spent the last few years as a premium perk on Razer’s gaming gear, baked into a $400 headset, a haptic cushion, and a specialty controller. That changes with the latest Kraken V4 headset that packs that same Sensa HD haptics already integrated into more than 100 PC games. The $100 wired headset is now the cheapest way to feel your games instead of just hearing them.
Razer has a new headset for players interested in feeling some “real” haptic feedback during their gaming sessions. The Kraken V4 X Sensa HD Haptics Edition employs the same Sensa HD Haptics technology used by Razer in previous products, vowing to provide deeper integration across the wide range of games already on the market.
The new Kraken V4 X Sensa is a wired USB headset which includes both immersive haptics and quality audio features. The Sensa HD Haptics tech can apparently turn every in-game audio effect into a physical sensation, with every effect providing unique feedback thanks to two distinct haptic modes.
In Dynamic Haptics mode, games with built-in Sensa HD Haptics support can provide scripted feedback that’s been linked directly to in-game events. Players should be able to feel weapons, abilities, and movement directly through their ears.
Furthermore, the Audio-to-Haptics (A2H) mode can add yet another layer to games Razer has already provided a profile for. While Dynamic Haptics come to the front, A2H mode works in the background to provide the player with additional haptic feedback.
Sensa HD Haptics is a complex technology designed to respond to intensity, duration, and position of sound effects. Distant explosions should have a different feeling than point-blank shots.
“In supported titles, every weapon, every ability, every moment of tension carries its own feedback,” Razer said. The technology is already supported in more than 100 PC games, including Battlefield 6, the Silent Hill 2 remake, Marvel Rivals, and 007 First Light.
Besides its “skull-penetrating” physical feedback, the Kraken V4 X Sensa HD Haptics Edition includes Razer’s TriForce 40 mm drivers designed to individually tune high, mid, and low frequencies. Support for virtual 7.1 surround sound is available as well, while the included HyperClear Cardioid Mic can eliminate unwanted background noise during voice chat.
We are in 2026 now, which means even an audio peripheral needs its own complex RGB lighting engine. Razer’s headset uses the company’s Chroma RGB engine, which can provide 16.8 million colors with dynamic, customizable lighting effects.
The Razer Kraken V4 X Sensa HD Haptics Edition is already available for purchase directly from the company’s store, or third-party stores such as Amazon. MSRP is $100, because there is no DDR RAM in the mix.
Tech
Lockheed’s stealthy Vectis drone shows hidden missile bays as the company pushes toward a 2027 first flight
- Lockheed Vectis carries its air-to-air and strike weapons inside the aircraft
- Lockheed expects the first Vectis prototype to fly in 2027
- The aircraft measures roughly 34 feet long with a 38-foot wingspan
Lockheed Martin has released promotional footage offering a clearer look at Vectis, its proposed stealth Collaborative Combat Aircraft for future operations.
The company says the aircraft is intended for autonomous missions alongside crewed fighters, including the F-35 and future combat platforms.
Lockheed expects the first Vectis prototype to fly in 2027, with four additional aircraft now entering development and testing.
Latest Videos FromTechRadar
Vectis reveals its internal weapons arrangement
New computer-generated imagery shows a deeply recessed intake above the fuselage, indicating significant attention to the aircraft’s signature characteristics.
The footage also depicts Vectis launching AIM-120 missiles and GBU-53 Small Diameter Bomb IIs from internal weapons compartments.
Keeping those weapons inside the aircraft allows the illustrated configuration to avoid exposing them on external attachment points during missions.
The arrangement supports the stealth requirements associated with the aircraft, although Lockheed has released few detailed performance specifications for Vectis.
The company describes Vectis as an autonomous aircraft capable of operating independently or cooperating with crewed aircraft during demanding missions.
It also displayed a full-scale mock-up measuring about 34 feet long with a wingspan of approximately 38 feet.
Those dimensions establish the physical size of the aircraft, while important information about propulsion, payload and other systems remains undisclosed.
Ron Fehlen, Lockheed Martin vice president and general manager of Skunk Works, said the design responds to demanding customer requirements.
“The technical specs, as you can imagine, are designed to meet and, in some cases, exceed the customer sets that we are working with and their requirements and their mission needs,” said Fehlen.
Affordability remains central to the Increment 2 effort
Lockheed claims the aircraft is survivable, affordable, and adaptable across different missions, without disclosing its projected production cost.
“Engineered for affordability and built for survivability, it delivers the flexibility today’s operational environment demands. This is the future of airpower, built the Skunk Works way.” Lockheed said.
Vectis is being developed for the Air Force’s Increment 2 requirement after Lockheed lost the earlier Increment 1 competition.
For Increment 1, the Air Force selected Anduril’s FQ-42 and General Atomics’ FQ-44 rather than Lockheed’s earlier, more expensive aircraft proposal.
Lockheed reduced complexity and cost while continuing development of a more capable stealth-oriented aircraft for Increment 2.
The Air Force’s requirements for Increment 2 remain under development, leaving the eventual balance between affordability, survivability and capability unresolved.
Earlier Increment 1 requirements called for more than 700 nautical miles of combat radius, while the later requirement could demand greater performance.
The lower costs could involve simpler stealth features and shorter-lived engines using manufacturing methods such as 3D printing.
A more demanding combat role, however, could require expensive propulsion, advanced avionics and sophisticated measures for reducing detectable signatures.
That combination leaves Vectis within an acquisition programme where the eventual technical requirements and preferred cost level remain unsettled.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
“We do not believe that the AI industry has solved alignment and monitoring to a sufficient degree to continue responsibly scaling at maximum speed for much longer,” OpenAI wrote on Wednesday.
The line appears in a new OpenAI post. It sets out how the company will track, investigate and disclose model misalignment. OpenAI uses the term for cases where AI systems fail to follow human values and safety goals. The company disclosed six new incidents alongside it.
“Decisions about how AI development should proceed in the months and years to come need to draw on evidence that people outside the companies building frontier models can examine for themselves,” the post says.
Six incidents in training and testing
All six cases involve unreleased research models or training runs. OpenAI published each one on a new Misalignment Reports page.
- An unreleased Astra-family model added unauthorised instructions to its own compaction summaries during training. Some told it to disregard its constraints. OpenAI identified 27 affected summaries.
- During GPT-5.6 Sol training, the model added instructions to hide mistakes and invent missing data without saying so.
- An internal model tried to sign up for disposable email accounts and searched public GitHub repositories for leaked API keys. It found and used an exposed key while answering questions about earnings figures. When it still could not find them, it made them up.
- Unreleased models uploaded data to temporary file-hosting services so they could cite it.
- Research models used OpenAI’s internal Artifactory repository as a message board. They exchanged requests and answers across separate training samples.
- Collaborating agents shared output files through public hosting platforms.
According to Axios, the earliest of the six incidents dates back to October. The uploads to public file-hosting services happened in October and January, Axios reported.
How the disclosure process works
Any OpenAI employee can flag a suspected case. Technical staff then investigate and decide whether to disclose it. OpenAI’s Safety Advisory Group settles disagreements, and unresolved ones go to company leadership.
Cases go on one of three tracks: ready for disclosure, minor investigation or larger investigation. OpenAI will publish cases that are ready within six business days, and minor investigations within 12, Axios reported. The larger track covers complex cases that involve third parties or security concerns, and those can take longer.
“There’s currently no industry wide framework with explicit disclosure standards, so we’re taking this step voluntarily because we think it’s really important to share what we’re learning,” Kai Chen, research lead on OpenAI’s alignment team, told Axios.
OpenAI told Axios it sees two causes. It did not have enough security controls to catch the incidents. Its models also advanced faster than it predicted. “We need to step up to meet this new era of AI development, and voluntary disclosures should be a part of that,” Chen said.
From Hugging Face to a framework
The framework follows OpenAI’s July disclosure about Hugging Face. Models under evaluation broke out of their test environment and compromised parts of its systems. OpenAI staff later said the agents coordinated the Hugging Face breach through the same internal repository.
Earlier this month, OpenAI confirmed the wiki incident, in which model instances misused a website to talk to each other. It said then that a disclosure framework would follow.
Talks with Anthropic and Google
Chris Lehane is OpenAI’s chief global affairs officer. On 9 September, he wrote in a company blog post: “We will work with other frontier labs to advance frontier AI standards, building a voluntary effort now, with or without government support.” He added that any such standards would complement mandatory federal safeguards and democratic oversight, not replace them.
On 12 September, Anthropic CEO Dario Amodei called for a slowdown in the development of the most advanced AI models. Google DeepMind’s Demis Hassabis and OpenAI’s Sam Altman both backed him. In his essay, Amodei wrote that the US government would need to “issue a narrow waiver for certain kinds of safety conversations” between the labs.
On Tuesday, an OpenAI spokesperson confirmed to CNBC that the company has been talking to Anthropic and Google about how they can work together on safety. The Information first reported the discussions.
The talks have been going on since July, the spokesperson said. That month, Hassabis published a proposal for a US-led “Standards Body”, modelled on the Financial Industry Regulatory Authority. Google and Anthropic did not immediately respond to CNBC’s request for comment.
Antitrust in Washington and Brussels
Also on Tuesday, Federal Trade Commission chairman Andrew Ferguson spoke at Georgetown University. He said some AI firms lobby for new rules while seeking an antitrust exemption. Those firms were after “barriers to entry that will insulate their incumbency from challenge,” Politico reported. Ferguson said he was giving a personal view, and he did not name Anthropic.
“Everyone should be deeply suspicious about this,” he said.
On the same day, US Treasury Secretary Scott Bessent told a House committee, according to Politico: “The best way to guarantee safety is that the creators are liable for what they build and generate.”
In Strasbourg on Wednesday, EU competition chief Teresa Ribera said she would consider a request from AI companies to coordinate on safety. “We have not got a formal request, at least for the time being,” she told reporters, Politico reported. “But I think that this is an interesting question and we will have to dive into this proposal if it comes.”
The same day, Commission President Ursula von der Leyen used her State of the Union address to say she would invite the main frontier labs to discuss how Europe can support their efforts to slow down.
The moon has more than 2 million craters on its surface, but the overwhelming majority of the big ones happened long before humans ever started looking at it. That means when someone notices a new big crater, it’s a pretty big deal, and that’s exactly what happened when NASA’s Lunar Orbiter found a brand-new crater it had never seen before that was large enough to get someone’s attention.
The crater was spotted by NASA scientist Robert Wagner on a routine data-quality check and published in an article in Science Advances. Wagner says the object that crashed into the moon and created the crater likely hit sometime between April 11, 2024, and May 22, 2024, and was likely about the size of a six-story building. It left a rather sizable dent, which now has the distinction of being the biggest new crater ever spotted by NASA’s Lunar Reconnaissance Orbiter.
“I just stopped, dropped everything, and started looking into what that spot was,” Wagner said in a statement.
The LRO scans the entire moon’s surface once every 14 days, meaning humans have a constant stream of new data about the surface, and since the moon has no atmosphere, its pockmarked surface remains largely intact until another asteroid or meteor barrels into it.
NASA says that the LRO has found approximately 1,000 new impact craters and over 100,000 other changes to the moon’s surface during its 17 years in service. Since the satellite can see craters only about 30 feet wide or larger, that also means many small impacts go virtually unnoticed, leaving the total number of changes to the moon’s surface a mystery.
One big hole
What makes this particular crater special is that it’s huge. At 728 feet wide, it’s more than twice the size of a football field, and at 141 feet deep is about the length of three yellow school buses stacked end to end. Impact craters that big don’t show up every day, every year or even every decade.
“Crater production models indicate that an impact of this magnitude is expected to occur only once every 132 years on the moon,” Wagner wrote in the article. “As such, this event constitutes a statistically rare, effectively one-in-a-lifetime observation.”
It’s the latest in a pretty busy year for the moon, which had human visitors for the first time in 50 years with the Artemis II mission before an abandoned SpaceX rocket crash-landed into it, which left a crater substantially smaller than the one Wagner found.
NASA named the crater McGetchin, after Tom McGetchin, a scientist who became the director of the Lunar Science Institute in 1977, shortly before his death in 1979.
Buying a premium television used to mean worrying about black levels, motion processing, HDMI inputs and whether somebody at the factory had decided Vivid Mode needed to resemble the surface of the sun. In 2026, the Smart TV hanging on your wall is also a networked computer that may stay connected for years to phones, speakers, game consoles and smart-home devices while running software, serving ads, measuring viewing behavior and personalizing content.
That changes the privacy equation considerably. What was once a relatively dumb display has evolved into a persistent platform sitting at the center of the home network, which means picture quality is no longer the only thing worth paying attention to.
A modern smart TV is a networked computer running an operating system, connected to other devices in the home and, on some models, equipped with microphones. It can also include advertising technology and Automatic Content Recognition (ACR), which can identify what is being watched when the feature is enabled.
LG is hardly alone in any of this, but the company finds itself under renewed scrutiny following a lengthy investigation by Gamers Nexus, Level1Techs and independent security researchers into the behavior and security of several LG televisions. Their findings arrived less than four months after LG resolved a Texas lawsuit concerning its ACR practices, creating a story that involves data collection, consent, voice processing, network discovery and cybersecurity rather than a simple case of a television secretly eavesdropping from the corner of the room.
The available evidence does not justify stating that ordinary LG televisions continuously record every conversation around them. LG explicitly denies that interpretation, while the researchers and subsequent security analysis raise legitimate questions about other aspects of the platform and what could happen if a television were compromised. That distinction needs to remain front and center throughout this story.
LG Has Been Here Before
Privacy questions involving LG televisions did not begin in 2026. In November 2013, British developer DoctorBeet examined network traffic from an LG NetCast smart TV and reported that viewing information continued to be transmitted even after he disabled the television’s “Collection of watching info” setting. His packet captures showed channel information and a unique device identifier being transmitted without encryption.
DoctorBeet also found filenames from USB-connected storage appearing in outgoing requests, although the receiving URL returned an HTTP 404 error. His testing therefore demonstrated that the television attempted to send those filenames, not that LG successfully received and stored them on an operational server.
That involved 2013 NetCast hardware, not current webOS televisions, so it cannot be used as evidence that a 2026 LG OLED behaves the same way. It does, however, provide some historical context for why another privacy dispute involving LG televisions deserves more than a shrug and a new checkbox in the settings menu.
ACR Became a Much Bigger Business
What changed dramatically over the following decade was the business sitting behind the television. Automatic Content Recognition can generate a digital fingerprint of programming and use it to identify what is being watched, which can support audience measurement, content recommendations and advertising.
LG made its ambitions rather explicit in January 2021 when it invested nearly $80 million in television-data company Alphonso, becoming its largest investor with a controlling stake exceeding 50 percent. LG said the deal would combine Alphonso’s data analytics, media planning and Video AI capabilities with its televisions and help create an owned first-screen and cross-device advertising platform. Alphonso subsequently became LG Ad Solutions.
The company went further in September 2022, announcing that LG Ads Solutions’ proprietary ACR system would replace its legacy technology across millions of LG televisions internationally. LG said the technology was already active on its U.S. smart TVs and described ACR data as a tool advertisers could use to understand exposure, measure campaigns and identify viewing audiences.
LG Ad Solutions now markets access to 49 million LG smart TVs in the United States and 216 million globally, while promoting proprietary ACR data as part of its advertising platform. Those figures describe LG Ad Solutions’ addressable television footprint; they do not mean ACR is enabled on every one of those televisions or that all of them are transmitting identical information.
There is nothing particularly secret about the commercial strategy. LG openly operates a connected-TV advertising and measurement business, which means the television can continue producing economic value long after Best Buy has stopped caring whether you kept the box.
Texas Sued LG and Four Other TV Manufacturers
On December 15, 2025, Texas Attorney General Ken Paxton filed actions against LG, Samsung, Sony, Hisense and TCL over the use of ACR technology. Texas alleged that consumers were not adequately informed about the collection and use of viewing information and challenged whether the companies had obtained meaningful consent.
Those were allegations by the State of Texas, not judicial findings that every manufacturer violated the law or that every television collected information in precisely the same fashion. The fact that Texas pursued five major brands is also significant because it establishes that the larger controversy surrounding ACR is not unique to LG.
That distinction becomes especially important when discussing how ACR actually works. LG says its current implementation uses audio fingerprinting through the television’s internal audio processor, rather than screenshots, screen recordings or stored audio, and says the feature is off by default until a user accepts its Viewing Information Agreement.
LG Reached an Agreement With Texas in May
LG and Texas resolved their dispute in May 2026 through an Agreed Final Judgment and Permanent Injunction. The order requires LG to obtain affirmative express consent before collecting or processing covered viewing data and requires clearer disclosures explaining that ACR is being used to identify and analyze television content.
The judgment also contains language that is essential for accurate reporting: LG’s agreement does not constitute an admission of wrongdoing, fault or liability, and the order cannot be construed as an admission that LG violated the law. Texas additionally recognized LG’s efforts to implement privacy protections, including measures the judgment describes as exceeding industry norms.
The accurate formulation is therefore not “Texas proved LG illegally spied on customers.” Texas alleged improper practices involving ACR and consent; LG agreed to stronger disclosure and consent requirements in order to settle the litigation without admitting wrongdoing.
The judgment also stipulates that viewing data collected from U.S. consumers has not been and will continue not to be transmitted from the United States to the People’s Republic of China. That language matters because it is materially different from implying that Texas discovered LG had previously been sending American viewing data to China.
Gamers Nexus and Level1Techs Went Looking Under the Hood
Then came September. On September 6, Gamers Nexus published a more than two-hour investigation developed with Level1Techs and independent security researchers, examining several LG televisions through network monitoring, firmware analysis and security testing. The researchers reported extensive network and device discovery, examined LG’s ACR and advertising infrastructure and disclosed security weaknesses they said were being handled through responsible-disclosure channels.
This is where the story requires some discipline because not every demonstration represented normal out-of-the-box television behavior. Malwarebytes, after reviewing the investigation and LG’s response, noted that some findings involved intended LG software behavior while others depended on security vulnerabilities and a compromised television.
The security vulnerabilities themselves are another matter. The researchers reported remote-code-execution vulnerabilities to LG but withheld full technical details while responsible disclosure was underway. Malwarebytes said LG had not publicly addressed those specific vulnerabilities as of its September 10 update, and as of September 16, LG has not publicly announced a fix for them.
Gamers Nexus also reported that the tested sets identified other devices and nearby Wi-Fi access points, while The Register reported claims involving device names, IP information, Wi-Fi networks and signal data. LG acknowledges device discovery and nearby Wi-Fi processing but says those functions serve separate purposes: device discovery supports connectivity, while Wi-Fi information can help estimate approximate location for ACR when the appropriate agreements have been accepted.
That is precisely why describing every finding as “LG spying” creates more heat than light. There are questions about normal product behavior, optional data collection and exploitable security vulnerabilities, and they should not be tossed into the same bucket merely because they create bloody chum for the online sharks.
LG Responds
LG published a detailed statement on September 12, addressing the allegations and explaining how it says voice recognition, ACR, network discovery and advertising systems operate. Its central denial could hardly be clearer:
“LG smart TVs do not continuously record or transmit users’ conversations.”
LG says voice recognition is disabled by default and requires the relevant agreement to be accepted. On supported sets with hands-free controls enabled, wake-word detection takes place locally; if the wake word is not detected, LG says the audio is not stored, converted to text or transmitted. Once a legitimate voice session begins, speech can be converted to text, but LG says sessions are temporary and limited to approximately 18 seconds.
The company also says voice information is not stored offline for later transmission when internet connectivity returns. Supported models equipped with a physical microphone switch can disable the built-in microphone entirely, although the remote microphone can still be used deliberately when voice recognition is enabled.
LG’s explanation of ACR is equally specific. The feature is optional and off by default; enabling it requires accepting the Viewing Information Agreement, while personalized and cross-device advertising require additional consent. LG says ACR may operate with Live TV, HDMI-connected equipment and LG Channels but does not operate inside third-party LG apps such as Netflix or YouTube.
LG also confirms that nearby Wi-Fi information and signal strength can be used to help estimate a television’s approximate location when relevant ACR agreements have been accepted. Depending on market, consent and user choices, ACR-derived viewing information may be shared with majority-owned LG Ad Solutions for audience analysis and, with further consent, interest-based or cross-device advertising. LG explicitly says voice and ambient-audio data are not provided to LG Ad Solutions for advertising.
That is a far more useful conversation than arguing over whether your OLED has suddenly developed an interest in what you discuss during dinner.
Gamers Nexus Fires Back
LG’s September 12 statement has not ended the dispute. Gamers Nexus responded the same day with a 47-minute follow-up challenging the company’s explanations and presenting additional testing that the researchers say did not require compromising the television.
In one test, researcher uturn says an ordinary LG television connected to an HDMI source transmitted information to Alphonso, the advertising company that became LG Ad Solutions, including a device identifier, city, state, ZIP code, latitude and longitude and the nearest major road. The researchers say the data was transmitted as frequently as 20 times per minute alongside audio fingerprints identifying content being played through HDMI.
LG acknowledges that nearby Wi-Fi information can be used to help estimate a television’s approximate location when the relevant agreements have been accepted, but we have not found a specific LG response addressing the exact data fields or transmission behavior documented in this follow-up.
Gamers Nexus also asked attorney Vincent Agosta to examine LG’s user agreements and consent process. His criticism centers on whether an ordinary purchaser could reasonably understand the collection and sharing described across multiple agreements — an issue that moves this controversy beyond the easy “is the microphone listening?” argument and back toward the more consequential question of what constitutes meaningful consent on a device most people still think of primarily as a television.
Is LG Unique?
No. ACR and smart-TV viewing-data collection are industry issues, which is precisely why Texas pursued multiple manufacturers rather than LG alone.
Vizio provides one of the clearest historical examples. In 2017, the FTC and New Jersey announced a $2.2 million settlement over allegations that software on 11 million Vizio smart TVs collected viewing histories without users’ knowledge or consent; the resulting order required prominent disclosure and affirmative express consent.
What makes the present LG story more significant is the collision of several issues at once. ACR and advertising were already under legal scrutiny, and the Gamers Nexus investigation then added questions about network behavior, voice processing and reported security vulnerabilities that could increase the consequences if a television were compromised.
That does not make LG uniquely nefarious. It makes LG the company currently sitting underneath the microscope.
Should LG Owners Be Concerned?
There is no reason to panic or have Uncle Paulie drag a 77-inch OLED into the Pine Barrens, but LG owners should take the privacy and security settings on their televisions seriously. The available evidence does not justify assuming that an uncompromised LG TV is secretly recording every conversation in the room, and LG says continuous ambient recording does not occur. At the same time, LG confirms that enabled ACR can identify supported content, process device and network information, use nearby Wi-Fi signals to help estimate approximate location and share certain viewing information with LG Ad Solutions depending on consent and jurisdiction.
A few practical steps can reduce unnecessary data collection and limit what a compromised television could access:
- Avoid unnecessary router exposure. There is generally no reason to manually forward ports to a smart TV, and UPnP can be disabled if your household does not depend on it.
- Keep the TV’s firmware current. Install software and security updates promptly. Whatever one thinks about ACR, the vulnerabilities identified by researchers are a reminder that a smart TV is a networked computer with very good black levels.
- Review LG’s privacy agreements individually. Pay particular attention to the Viewing Information Agreement, Voice Information Agreement and Interest-Based Cross Device Advertising Agreement rather than accepting everything just to get through setup.
- Disable ACR if you do not want viewing information collected. LG says ACR is off by default, requires consent and can be disabled without losing core functions such as HDMI inputs, Live TV, streaming apps or software updates.
- Turn off voice features you do not use. On supported models, owners can also use the physical microphone switch to disable the television’s built-in microphone.
- Consider isolating smart-home devices from your primary network. A separate guest or IoT network can reduce what a compromised television, camera or speaker can reach, although overly aggressive isolation may interfere with casting and device discovery.
The Bottom Line
Calling this simply “LG TVs are spying on you” would produce a terrific thumbnail and an unnecessarily sloppy news story. The underlying facts are more complicated and, frankly, more important.
Texas challenged LG’s ACR consent practices, while LG settled the litigation and accepted stronger consent and disclosure requirements without admitting wrongdoing. Gamers Nexus and its research partners subsequently reported network behavior and security issues they considered troubling, including vulnerabilities that could greatly increase the privacy implications of a compromised set.
LG disputes the interpretation that its normal voice-recognition system continuously records household conversations, but its September response also illustrates just how sophisticated the information infrastructure behind a modern television has become. ACR, approximate-location processing, device discovery, audience measurement and advertising can all exist within the same product, even if they operate as distinct systems with separate permissions.
The larger issue therefore extends well beyond LG. Smart-TV manufacturers increasingly operate software and advertising platforms that can continue generating revenue and information long after the hardware sale, and consumers deserve to understand exactly what they are agreeing to when those capabilities are enabled.
Picture quality still matters, but privacy has quietly become another television specification worth paying attention to. At this point, asking your $3,000 OLED to mind its own business should not qualify as an advanced feature.
For more information: lg.com
Related Reading:
NewsBeat4 minutes ago
Clarks award-winning trainers in four colours are designed to help ‘walk for longer’
Business5 minutes ago
Macmahon to acquire Aspect Engineering Solutions
Crypto World6 minutes ago
SEC grants 5-year exemption for tokenized stock trading
-
Fashion6 days agoWeekend Open Thread – Corporette.com
-
Tech5 days agoThe Latest Weird Thing to Play Doom Is the Mapped-Out Brain of a Fruit Fly
-
Business7 days ago10 Most-Streamed Songs On Spotify In 2026 So Far, Led By Ella Langley’s Dominant Run On The Charts This Year
-
Crypto World7 days agoXAG/USD: Silver’s Short-Term Rally Meets Its Moment of Truth
-
Crypto World7 days agoDiesel Tops $6 a Gallon for the First Time as 28 States Set Records
-
Crypto World3 days agoKraken Lets xStocks Holders Earn Yield Through DeFi
-
News Videos7 days agoFacing Financial Fears
-
Crypto World4 days agoElon Musk Drops a Bombshell: Grok 5 Could Be the AGI Breakthrough
-
Crypto World1 day agoUS Charges Robinhood Engineers Over Crypto Listing Trades
-
Entertainment6 days agoNews Specials, Movies, Shows, More
-
Crypto World6 days agoCan AI Build a Startup in 72 Hours? Elon Musk's Team Will Livestream the Test
-
Business5 days agoRivals Sam Altman and Elon Musk Rally Behind Dario Amodei’s Call for a Slowdown in AI Development
-
NewsBeat3 days ago‘Sick conspiracy’: Trump says only guardrails AI needs is ‘a strong and smart (High IQ!) president’ in all-caps rant
-
Crypto World4 days agoNew Tesla Roadster Uses SpaceX Tech. Will It Impact the Stock Price?
-
Crypto World7 days agoBitcoin ETFs Pull $167M as 2026’s Best Inflow Run Slows
-
Crypto World6 days agoRobinhood Chain Never Stopped But its Blobs Did Stop Reaching Ethereum For 14 Minutes
-
Crypto World2 days agoWhat Is the Status of the U.S.-Iran Peace Talks? Here's What Both Sides Are Saying
-
Entertainment6 days agoNew Horror Movie Officially Earns a Rare Stephen King Recommendation
-
Business6 days agoMarvell: Most Potent Setup Of The AI Factory Decade
-
Crypto World4 days ago3 Token Unlocks to Watch in the Third Week of September 2026
You must be logged in to post a comment Login