Listen Labs, a market research startup that uses voice AI to conduct customer interviews, recently signed a term sheet for a $125 million Series C at a $1.5 billion valuation, with Menlo Ventures set to lead the round, according to several people with knowledge of the matter.
But that round never closed, the people said. Listen Labs walked away from the signed term sheet, a rare occurrence in the venture world and one that is generally frowned upon, according to VCs.
The financing likely collapsed because of acquisition talks with Salesforce. The CRM giant has recently held talks to buy Listen Labs for around $2 billion, Business Insider reported. The discussions are not finalized, however, and may not result in a deal, the outlet notes.
Listen Labs is one of the leading startups in the rapidly growing field of automating customer research with AI. The three-year-old startup has about $30 million in annualized revenue, about three times more than Simile, a competing startup that predicts human behavior, according to two people familiar with the companies’ financials. In late July, Simile announced that it had closed a $200 million Series B at a $2 billion valuation led by Greenoaks — likely setting a new valuation benchmark for Listen Labs, one person said.
If talks with Salesforce collapse, several VCs told TechCrunch that they expect Listen Labs to return to market and target a valuation of $2 billion or higher.
While acquiring Listen Labs could strengthen Salesforce’s AI capabilities by using the startup’s AI to help predict customer needs, the CRM giant may ultimately decide that paying a 67 times revenue multiple is too steep a valuation, according to a person with experience negotiating exits to Salesforce.
Listen Labs, Salesforce, Menlo Ventures, and Simile did not immediately respond to requests for comment.
Listen Labs was co-founded in 2023 by Florian Jüngermann, a former German national champion in competitive computer programming, and Alfred Wahlforss, who previously founded a staffing startup called Bemlo. The two met while pursuing master’s degrees at Harvard.
Listen Labs’ AI develops survey questions and interviews customers over audio or video. The resulting conversations are then packaged into reports and PowerPoint presentations, similar to those traditionally produced by human market researchers.
Fortune 500 companies rely on this type of research to gauge customer needs and satisfaction with their brands and products, but traditional market research is expensive and can take weeks to complete.
Listen Labs’ technology helps reduce the time and cost of these projects, enabling companies to quickly understand how customers are reacting to product changes and iterate on them more efficiently.
The startup’s customers include Microsoft, Canva, Anthropic, and Sweetgreen. Listen Labs and Simile aren’t the only startups using AI to disrupt the customer research market.
Besides Simile, competitors in the space include Outset, Keplar, and Aaru. While some platforms automate interviews with real humans, others startups — like Aaru and Simile — take a synthetic approach, using AI to simulate human behavior and predict responses without interviewing anyone at all.
Listen Labs was previously valued at $500 million when it announced a $69 million Series B round in late January led by Ribbit Capital, with participation from returning backers Sequoia, Conviction, and Pear VC.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Amazon named cybersecurity veteran Kevin Mandia to its board of directors, adding new security expertise a few months after former NSA director Keith Alexander stepped down.
Mandia founded Mandiant, the breach-investigation firm Google acquired for $5.4 billion in 2022, and remained at the search giant as a strategic advisor through July 2025, according to his LinkedIn profile. He now leads Armadin, an AI security startup he started last year.
Amazon said in its announcement that “cybersecurity is one of the most consequential risks and responsibilities organizations face today, and the threat landscape continues to evolve rapidly alongside advances in AI.”
Amazon added a cybersecurity specialist to its board in 2020, when it elected Alexander, who also led U.S. Cyber Command. Mandia comes from the other side of the field, with two decades spent investigating corporate breaches rather than defending government networks.
His appointment also puts an AI security entrepreneur on the board of a company whose cloud infrastructure underpins much of the internet. Armadin, founded in September 2025, uses AI to run attacks against corporate networks, probing defenses the way an intruder would.
The board’s Security Committee, which oversees Amazon’s cybersecurity policies and its response to significant cyber incidents, is now chaired by Dan Huttenlocher, dean of the MIT Schwarzman College of Computing. Mandia joins as a member, along with former Bridgewater co-CEO Jon Rubinstein.
Amazon also named Mandia to the board’s Audit Committee, according to a securities filing.
Mandia received 4,086 restricted stock units in connection with his election to the board, vesting in three equal annual installments beginning Nov. 15, 2027, the filing shows. The shares were worth about $1.03 million at Amazon’s closing price Wednesday.
The filing disclosed that his sister-in-law, Kristin Mandia, is an Amazon employee with an annual salary of $185,000. The company said her compensation is consistent with that of other employees at her level with similar responsibilities.
Oil is your car’s life blood. From the engine to the transmission to the differential — anywhere with the potential for metal-on-metal contact, you’ll need some sort of lubricant to keep it from overheating. Every type of oil is designed for different tasks as well; while some may be theoretically interchangeable, they often have different additives that make them better suited for one task or another. Take hydraulic oil versus motor oil, for instance. Both have a lot of the same qualities, but the former is best suited to accepting high pressure and the latter high thermal loads.
The same thing applies here with hypoid gear oil, such as many API GL-5 oils, versus regular gear oil. In short, hypoid gear oil is meant to be used with steel gears under the pressures inherent to hypoid gearsets, as opposed to regular gearsets.
Generally, gears use the same principle whether it’s a transmission or a windmill; you’re taking two shafts and meshing them together to transfer torque. The difference comes in how those gears interact. You have shafts that can go laterally or extend sideways, some that are articulated and others that are static. Ring gears, planetary gears, straight-cut or beveled — all these produce different pressures that require different oils, hence why manual transmission fluid (MTF) and automatic transmission fluid (ATF) are different.
But what’s special about hypoid gears that you need hypoid oil? Hypoid differentials are arranged differently; you have an input shaft that’s lower or higher than the center of the output shaft gear’s center, a setup which has its own benefits but also requires specialized lubricants because of the friction involved in laterally sliding two metal surfaces together.
Yes, car axles are complicated, so let’s address this first: What is a hypoid gear? To better visualize how this works, ball one hand up into a fist, then place your index finger beside it. Imagine your fist is a gear, and your index finger is an input shaft. That input shaft is spinning, and as it spins, it turns the bigger gear called a ring gear — your fist, in this case, which is how a driveshaft turns your car’s wheel in a typical rear-wheel drive car.
Let’s assume that the input shaft is perfectly dead-center where the output’s axis of rotation is. Your finger has a certain amount touching your fist in that case; that’s where your car’s input shaft contacts the output shaft. Okay, now for the real trick — rotate your finger so that your whole index finger is touching your fist. You have way more surface area between the two; that’s what a hypoid gear is. You’re putting the input shaft in a place where more of it is in contact with the ring gear on the output shaft.
Now, when you think about it, you can’t just place the two gears offset to one another and have them be cut perfectly straight. It’d be like moving your jaw to one side and trying to close your mouth. The teeth wouldn’t be able to mesh, so instead you cut the teeth in a spiral pattern and cut the ring gear with a curve to its teeth. That way, as the gear rotates, it’ll mesh without any friction issues because there’s minimal lateral resistance (like grinding your teeth).
Any gearbox needs fluid, as we’ve established, and hypoid gears are no exception. You’ll frequently see hypoid gears running with API GL-5, but what does that even mean and can you use regular gear oil instead? In short, no. Hypoid gear oil has specialized additives which are toleranced for the pressures placed upon them by the spiraling motion of the beveled gears.
A lot of regular manual gearboxes can also use this fluid as well; look on the back of any bottle of GL-5 and it’ll likely say it’s rated for use in them. Most modern gearboxes use steel gears, which GL-5 protects by clinging onto the surface and acting as a sacrificial barrier. The oil prevents the metal surfaces from generating friction, therefore helping your transmission last longer. The main difference is the additives. GL-5 has specific sulphur/phosphorous additives that cling onto metal better under pressure, which is why they’re so good with hypoid gears. The downside is that this additive is corrosive to brass, so if you have a car with brass synchronizers in the gears, you have to use GL-4 instead — something I have to do personally, as my car has brass synchros.
As for using other gear oils with hypoid gears, you technically can — again, much like using motor oil in hydraulic lines. But it’s not a good idea; the point of GL-5 is that it clings to metal surfaces. Without that action, your differential will struggle to lubricate itself, leading to the gears wearing out. Signs of differential failure include ticking or clunking noises when driving, a generally noisy ride coming from the axle, or odd behavior in tight corners — so don’t forget to change your diff oil.
Cisco has confirmed that a maximum-severity authentication bypass vulnerability tracked as CVE-2026-20079 in its Secure Firewall Management Center (FMC) software is being actively exploited in attacks.
The vulnerability has a maximum CVSS score of 10.0 and allows unauthenticated, remote attackers to bypass authentication and execute scripts and commands as root on vulnerable devices.
“In August 2026, the Cisco PSIRT became aware of active exploitation of this vulnerability,” Cisco updated its CVE-2026-20079 advisory to say on Wednesday.
Cisco did not disclose when the attacks began, who was behind them, or what post-exploitation activity was observed.
Cisco first disclosed CVE-2026-20079 in March, when the company said it had no evidence that the vulnerability was being exploited in attacks.
The flaw is caused by an improper system process created at boot time and can be exploited by sending crafted HTTP requests to the web interface of an affected device.
A successful attack can allow an unauthenticated attacker to execute scripts and commands on the device with root privileges.
The vulnerability affects Cisco Secure FMC Software and Cisco Security Cloud Control Firewall Management. Cisco says it has already patched the cloud-hosted Security Cloud Control service.
Cisco says there are no workarounds and recommends that customers upgrade to the latest software release.
Today, the U.S. Cybersecurity and Infrastructure Security Agency (CISA) added CVE-2026-20079 to its Known Exploited Vulnerabilities (KEV) catalog, ordering Federal Civilian Executive Branch agencies to secure vulnerable systems by September 12, 2026.
While Cisco says its security team became aware of active exploitation of CVE-2026-20079 in August, IOCs published in a July advisory update suggest the flaw may have been exploited earlier.
On July 29, Cisco disclosed another Secure FMC vulnerability, tracked as CVE-2026-20316, caused by static credentials for a low-privileged account.
Cisco said at the time that CVE-2026-20316 had been actively exploited in attacks and assigned it a High severity rating because the access could be combined with other Secure FMC vulnerabilities to elevate privileges.
As BleepingComputer reported at the time, Cisco also updated the CVE-2026-20079 advisory to include the same indicators as CVE-2026-20316, but did not confirm the flaw was exploited.
Cisco told administrators to search /var/log/messages for activity related to /var/tmp/license.tmp and shared the following example log entry:
Jul 23 16:16:33 firepower sudo: www : PWD=/ ; USER=root ; COMMAND=/usr/local/sf/bin/package_info.pl /var/tmp/license.tmp --lsm
Cisco says that if this entry is found, the vulnerability “may have been exploited” on the examined Secure FMC device.
The example log entry is dated July 23, weeks before Cisco says PSIRT became aware of exploitation of CVE-2026-20079 in August.
Cisco also released the same Secure FMC hot fixes for both CVE-2026-20316 and CVE-2026-20079.
At the time, BleepingComputer contacted Cisco to ask whether the two vulnerabilities were connected, whether CVE-2026-20079 had also been exploited, and whether Cisco intentionally added the shared indicator to both advisories.
Cisco did not answer the questions directly and instead shared the following statement:
“On July 29, 2026, Cisco released software fixes to address vulnerabilities in Cisco Secure Firewall Management Center (FMC). Details are outlined in the security advisories (Static Credential vulnerability, Authentication Bypass vulnerability), and Cisco strongly recommends customers immediately apply the available fixes,” a Cisco spokesperson told BleepingComputer.
“Customers needing support should contact the Cisco Technical Assistance Center (TAC).”
Cisco’s latest update now confirms that CVE-2026-20079 has been exploited, but does not clarify whether the July 23 activity included exploitation of both vulnerabilities.
However, the same IOCs for both flaws, identical July hot fixes, and the July 23 log entry suggest both vulnerabilities may have been used in the same attacks.
Cisco advises customers who discover the indicators of compromise to contact its TAC for support, warning that installing the hot fixes will prevent future exploitation but will not remediate devices already compromised.
Overall prevention scores can hide what happens after initial access. Once attackers are using valid credentials, prevention drops sharply.
The Blue Report 2026 measures defenses technique by technique across 338 million simulations run in customer production environments.
ai and ml
Claude’s Felony Bench rap sheet is now as long as OpenAI’s
Amid industry soul-searching¹ about the possibility of AI improving itself to the point that it kills everyone, Anthropic has revealed yet another incident that would qualify as a crime if perpetrated by a person.
The AI biz published “an alignment assessment” detailing four times Claude models accessed third-party systems without authorization.
The company has already reported three of the incidents. Evidence of the fourth was lurking in a session transcript dating back to January 2026 when the misbehavior occurred.
Anthropic found the first three by scanning around 141,000 transcripts where Claude could have obtained internet access during evaluation. It missed the fourth initially because “our scan relied on an agentic search.”
Felony Bench, a tongue-in-cheek record of cyber intrusions carried out by major AI companies without consequences, has added this newly-discovered incident to its rap sheet of rogue AI actions.
The January 2026 AI trespass involved an early version of Claude Opus 4.6, which was given a Capture the Flag (CTF) challenge under the oversight of the third-party model evaluator where the other hacking events occurred.
Opus 4.6 managed to sabotage its chances of success by disabling the machine it was targeting. It assigned the device an IP address that already existed on another piece of hardware, rendering the target unreachable and making it impossible to solve the challenge.
Those familiar with other incidents where AI models violated third-party systems may recall that unsolvable tasks represent a common catalyst for misbehavior. Models exhaust all aligned options, and then turn to transgressive approaches.
Opus 4.6 might have been an exception, but when it tried to abort the task after recognizing that it could not reach the target machine, it failed to do so “due to a misconfiguration in [the model’s] evaluation harness.” It failed to shut down not just once but seven times.
So it continued onward, trying other expected means to reach the target machine but failing. Then it explored further.
“The model discovered a machine belonging to a third party that it was able to access, and stated that it believed this third party was part of the CTF,” Anthropic explained in its post. “Inside the machine, the model found a file listing a password, which it used to gain admin access to the system.”
The model went on to gather more credentials, and modified a system setting to make it easier to access the personal information of an individual associated with the third party evaluation organization.
Opus 4.6 might have done more but for the fact that it exhausted its token budget, bringing the session to an end.
Anthropic says it’s not as concerned about this incident as the others because the model tried to abort its task.
“While the model’s disregard for the possibility that it might be harming real systems or people is concerning, many of the behaviors described here have changed considerably as our training has evolved across model generations,” the company said.
Anthropic said it considers these incidents serious but expects current training approaches “are likely able to address the specific alignment failure modes observed in these incidents.”
And if company training methods fall short, there’s no real consequence to anyone at Anthropic other than writing up a revised alignment assessment. ®
¹ The term “soul-searching” is figurative and is not intended to indicate a belief that the technology industry has a soul.
Connecting the dots: Tesla’s Cybercab may not have a steering wheel or pedals, but there may still be a way to drive it manually. A video posted on X shows an unusual in-car display setup: the touchscreen showing a virtual joystick that lets someone drive the robotaxi by dragging a finger across the screen. “When I got in the vehicle, the display was in a different configuration with a virtual joystick for manual driving,” a rider wrote, adding that he had “a feeling I wasn’t supposed to see this” before trying the feature out anyway.
The Cybercab runs on a single central touchscreen. In the video (below), it’s split in two: a control panel on the left, the vehicle’s camera view on the right. The left side includes the joystick, a prompt to “Close cabin doors,” door controls, and a button to call support.
The right side shows the car’s surround-camera feed while parked in what looks like a lot, but not all of the imagery appears to be live. The side-camera views seem to show a previous location near a Supercharger rather than wherever the car is actually sitting.
– Matthew Skrzypczak (@MatthewSkrzypc1) September 7, 2026
Tesla hasn’t publicly shown or discussed this manual-driving screen. Still, the feature addresses a basic problem with a vehicle built without conventional controls: even a robotaxi occasionally needs to be moved by a person, whether that’s at a depot, in a lot, or when it’s simply blocked in. The on-screen joystick appears to be Tesla’s answer – rather than adding a steering wheel and pedals, the company built a touch-based control option straight into the software.
Tesla has positioned the Cybercab as a purpose-built autonomous vehicle, stripping out the equipment normally reserved for a human driver – steering wheel, accelerator, brake pedal. The joystick interface suggests the company has kept a limited way to take control when needed after all.
The feature surfaced just days after Tesla launched the Cybercab in Austin on September 3. That same day, the NHTSA opened an audit into how Tesla certified the vehicle without a steering wheel, pedals, or mirrors. The audit covers roughly 1,000 Cybercabs, and NHTSA is examining whether Tesla was correct to determine that certain federal safety standards simply don’t apply to a driverless vehicle.
Tesla’s argument rests on the idea that a fully autonomous vehicle doesn’t need equipment built for a human driver. The joystick screen doesn’t necessarily undercut that case – it may be meant only for limited manual repositioning rather than everyday driving.
Still, it’s a reminder that the Cybercab isn’t entirely beyond human control.
Tesla chair Robyn Denholm has hinted as much before. “If we have to have a steering wheel, it can have a steering wheel and pedals,” she said.
Anker Prime 26K 300W, priced at $179.98 (was $230), arrives at a useful moment. New iPhones (18 Pro, 18 Pro Max, Duo, etc.) want faster wired charging, and plenty of laptops still expect a wall brick that can push 100W or more. This pack tries to cover both without forcing you to haul a second charger.
Model A110A has a battery capacity of 26,250mAh, or 99.75Wh, which falls short of the 100Wh limit for airline carry-on size, but is still close (yet fully legal). The dimensions are 6.3 by 2.5 by 1.5 inches, with a weight of around 1.32 pounds (600 grams). To be honest, this pack is more suitable for a backpack than a jeans pocket. The two USB-C ports can deliver 140W each using USB Power Delivery 3.1. Yes, that includes a 28V 5A step, which high-wattage laptops actually use, as well as a USB-A port capable of 22.5W (or 20W if all three ports are in operation at the same time), for a total of 300W. Although 300W is the maximum you’d receive if everything was maxed out at the same time, in fact you can easily run two hungry laptops at 140W each as the phone drink away at port A.
Sale
Anker’s lab results reveal that an M4 Pro MacBook Pro can reach 50% in 27 minutes, while an iPhone 16 Pro does so in 26 minutes. Independent tests confirm this, with an iPhone 17 Pro Max going from 1 to 50% in just 22 minutes, leaving just enough energy for about three full phone charges. Ever notice how some computers aren’t fully charged? A 14-inch MacBook Pro reached 100% battery life and still had some power remaining. What’s preventing the rest? An M16 has a fairly conventional stop of 70-80%, but the conversion heat consumes a portion of the stored energy.
The second half of the story is recharging speed, with each USB-C capable of taking up to 140W. Both of those C ports can handle up to 250W if you use 5A cables, ideally in conjunction with two powerful wall chargers. Anker claims to be able to achieve 40% in 10 minutes and 50% in 13 minutes. Testers who used dual input reported a complete refill in 48 minutes to slightly more than an hour, depending on the wall chargers. You can also pass-through charging, which allows you to charge your phone(s) while the pack charges itself. The box contents, however, are rather basic: a 3.3 foot 240W cable, a travel pouch, and a simple tutorial to get you started.
The front of the pack has a color screen that rotates with the pack (so you don’t have to twist it back and forth in your hand to check on it), which displays the remaining percentage, power on each port, power input, and battery temperature. Anker also offers a phone app with additional capabilities such as volt, amp, and watt graphs, as well as firmware upgrades and smart allocation modes that lock power to a single port. PowerIQ 4.0 does its best to decide how to distribute energy across numerous devices, and ActiveShield 4.0 is quite concerned about temperature; tests rarely saw the pack get beyond 41°C in a typical room, though a lengthy 140W session will still warm the shell. Just a quick remark on the display side: due to its gloss finish, it is a bit of a fingerprint magnet, and there is no built-in cord, wireless pad, or speaker to help you find a lost pack.
Some of its on-paper specification isn’t especially encouraging, but spend some time in the company of the Yamaha NX-70A and you’ll discover that it’s not how big your numbers are but what you do with them that’s important. When it comes to mature, well-realised sound there’s nothing at all sub-optimal here…
Balanced, insightful and periodically absorbing sound
Good control options include a very thorough app
Excellent standard of build and finish complements the mildly interesting aesthetic
Not as hi-res as any number of alternatives
Another input or two wouldn’t go amiss
Accompanying stands are not cheap
Review Price:
£2599
Power
320 watts of Class D amplification
Audio
32-bit/384kHz PCM / DSD256
Speakers
Harmonious Diaphragm driver technology
Where its audio division is concerned at least, Yamaha is not the sort of company to rush things – the NX-N500 wireless stereo speaker system launched a decade ago, it’s true, but since then it’s been concentrating its efforts elsewhere.
Which means it’s had 10 years and more to keen over the NX-70A – has it been worth the considerable wait?
We’re hardly in KEF LS Luxe territory where design is concerned, but nevertheless the Yamaha NX-70A are mildly unusual lookers and mildly interesting as a result. There’s nothing odd about the cabinet dimensions – but the gently curved front baffle and the rather more aggressively curved rear serve to make the speaker look softly triangular when viewed from above.
There are no parallel lines here, and this of course is handy when it comes to minimising internal standing waves. View the cabinet from the dead-ahead, meanwhile, and it becomes apparent that the rear of the cabinet is taller than the front – so the 5mm thick aluminium top plate slopes forward somewhat.
Despite the relatively complex nature of the cabinet design, though, gratuitous design flourishes are few and far between. No matter if you go with the black or white option for finish, there are copper-coloured accents visible – the horseshoe-shaped heatsink at the rear of the top-plate, the manufacturer badge on the front of the top-plate, the input/output plates on the rear of the speaker, and the brightwork surrounding each driver, are all in this colour.
As has long been the case with Yamaha, the standard of build and finish that’s in evidence here is basically impeccable. The wooden cabinets (and their aluminium top-plates) are screwed together flawlessly, and are finished to a very high standard. The exposed screw-heads on the top-plate won’t be to everyone’s taste, mind you.
Yamaha is not alone in offering bespoke stands for its wireless stereo speaker system, but the fact the SPS-70A stands cost around £500 per pair means they’re merely expensive rather than extortionate (and, again, it means we’re not in KEF LS Luxe territory). The speakers bolt to a top-plate that’s exactly the same shape as the foot of the speaker, and there’s some cable-management in the pillar too. Soft or pointed floor spikes are provided.
Controlling the NX-70A can be done in a couple of different ways. There are no controls on the speakers themselves beyond a few buttons on the rear of the primary speaker to initiate Bluetooth pairing, wireless connection to the secondary speaker, let it know if it’s the left or right channel and what-have-you.
The primary speaker has a display panel at the bottom of the front baffle that let you know what’s what as regards volume level, the tune that’s playing and so on (the secondary speaker has a matching panel but it’s not a display – it simply has a tell-tale to let you know it’s powered on).
And the supplied remote control handset looks and feels nice enough, and gives you control of volume, power, input selection and the basics of playback.
Don’t imagine Yamaha has been sitting on its hands while its most obvious rivals all stole a march in the premium wireless stereo speaker system market. Better late than never, after all – especially if it’s the development of a selection of racy-sounding technologies that have held you up all this time.
To start at the end: each NX-70A uses a 30mm dome tweeter and a 130mm mid/bass driver, bolstered by a rear-venting bass reflex port, to deliver sound. That’s a big tweeter, sure – but even more unusually, the tweeter and the mid/bass driver are made of identical material.
Harmonious Diaphragm is a blend of Zylon (a synthetic polymer fibre) and spruce wood. The idea of using the same material across both drive units is to be able to offer absolute consistency of tone over the entire frequency range.
Each tweeter is in receipt of 60 watts of Class D power, while each mid/bass driver gets 100 Class D watts to deal with – an all-in total of 320 watts of amplification really ought to prove adequate in any realistic scenario.
In an effort to enhance sound quality, Yamaha has developed the Synergistic Drive unified circuit design, which sees each drive unit connected directly to the amplifier circuit. By being better able to manage the flow of the electrical current using this technology, goes the theory, the distortion inherent in traditional amplifier/speaker interactions is minimised.
Naturally it is critical to the process that the drivers and the amplification are developed simultaneously as a dedicated pair. The upshot of all of this extremely thorough engineering is, according to Yamaha, a frequency response of 50Hz – 35kHz (when the speakers are wired together – it modulates to 50Hz – 21kHz if you prefer a wireless connection).
Yamaha Parametric Room Acoustic Optimiser (or YPAO to use its only slightly snappier abbreviation) is the company’s proprietary automatic sound optimisation technology, and is accessed via a bundled microphone that measures the characteristics of the room the speakers are positioned in.
There’s a socket on the rear of the primary speaker into which the mic can be connected, and it’s no kind of hardship to run the YPAO routine using the MusicCast control app.
The app allows access to each of the available sources and also offers playback and volume control. There are a number of EQ adjustment options available too. There are 40 (!) slots for Favourites (which can be playlists, songs, radio stations or whatever you fancy, really), the ability to build and access a quick playlist from various sources, create a number of routines (for instance, all power off) that can be accessed quickly and easily.
And then there are further possibilities if you’re using other MusicCast equipment – building a surround-sound or multiroom system are the most attention-grabbing. The overall layout of the app is a little busy and crowded, but there’s no disputing its thoroughness or its stability.
The primary speaker also includes all wired and wireless connectivity. Bluetooth 5.3 is on board, and it’s compatible with the plain vanilla SBC and AAC codecs. Dual-band Wi-Fi is also available, and this means the system is compatible with Apple AirPlay, Google Cast, Spotify Connect and TIDAL Connect.
In addition, the NX-70A is Roon Ready and internet radio is available too, along with access to a podcasts platform and DLNA-compliant storage on a common network.
You may prefer to use Ethernet, of course, and there’s the relevant socket as part of the suite of physical connections. It lines up alongside an HDMI eARC socket, a digital optical input, a USB-A slot and a 3.5mm line-level analogue input.
Outputs run to a pre- for a subwoofer and an RJ45 for making a physical connection to the secondary speaker (the secondary speaker has a corresponding RJ45 and a mains connection – and that’s the lot). By prevailing standards there’s not a lavish amount of connectivity here, but nevertheless the NX-70A will support a fairly extensive system.
Any and all incoming digital audio information is pored over by an ESS ES9017S DAC chipset that’s capable of dealing with files of up to 32-bit/384kHz PCM and DSD256 resolution. If the speakers are wired together, everything is up- or downscaled to 24-bit/96kHz resolution, and this drops to a reasonably healthy 24-bit/48kHz resolution if they’re connected wirelessly.
If you came for shock and awe, Yamaha is probably not all that unhappy to disappoint you. There’s nothing showy or gratuitously attention-grabbing about the NX-70A – its many charms are altogether more grown-up than that.
That’s not to say it’s short of vigour, or that it is unwilling or unable to attack those recordings that require it. But this is a balanced and poised system, and it’s not prone to bluster.
Once through a 24bit/192kHz FLAC file of Nina Simone’s reading of Pirate Jenny (downscaled by the Yamaha, of course) reveals a lot of what the NX-70A is about. Its approach to the frequency is very nicely balanced, with no area over- or underplayed and due attention paid at every stage.
Low frequencies are deep, solid, properly shaped and alive with variation – details of texture and timbre are readily available. The top of the frequency range is a nice combination of bite and substance – there’s a kind of benign assertiveness to the way the Yamaha delivers treble sounds that suits its overall attitude of quite polite authority.
In between, the NX-70A delivers that inimitable vocal in the most eloquent and revealing fashion. Detail levels are sky-high, with a huge amount of information regarding character and attitude (both peerless) made available as readily as the stuff around tone and timbre. Simone is possessed by the character she’s singing about during the course of this recording, and you won’t hear that fact communicated any more readily by any even remotely price-comparable alternative.
The tonal balance is just as judicious. The Yamaha is almost fanatically neutral in this respect, seeming to add or subtract nothing where temperature is concerned – any suggestion of heat or hint of chill in a recording will almost certainly be inherent in the recording or the work of the source.
Soundstaging is expansive, coherent and controlled. There’s a fair amount of free space between the elements of a recording, so each one gets elbow-room in which to operate – but at the same time there’s a reasonable sensation of unity and singularity to the presentation. The sound has no problem escaping the left/right confines of the cabinet, and there’s a fair suggestion of height to the stage too – but what’s most impressive is the unequivocal way in which every element is locked in position and secure in its space.
Switching to a vinyl copy of Julian Cope’s Peggy Suicide allows the NX-70A to demonstrate its talent with dynamic variations of all kinds. It breathes deeply enough to make the upshifts in intensity and volume both obvious and significant, and it pays sufficient attention to allow harmonic variations to be apparent too.
And this recording also allows the system to showcase its talent with rhythmic expression – low-end information is tightly controlled at the point of attack, and there’s real confidence to the way rhythms are described as a result.
It also allows the Yamaha to reveal a decent capacity for drive and attack, too. As I started by saying, this is not the sort of system that’s going to grab you by the lapels – but neither is it undemonstrative or in any way passive. The more vigorous the content you’re playing, the more direct and positive this system sounds – but at no point does it ever threaten to lose the run of itself.
Rigorously balanced and articulate sound from a two-speaker system with the look to match the sound seems like your sort of thing.
You’re expecting sonic fireworks no matter the sort of music you’re listening to.
Instant gratification is all well and good, but sometimes a slower burn eventually leads to a more rewarding experience.
The NX-70A is not the sort of system that’s going to sell itself to you in the first few seconds – but its charms are of the enduring kind, and it’s worth allowing a little longer than you might be expecting to let it reveal what it’s capable of.
I bolt the NX-70A to the SPS-70A stands and plug both speakers into the mains. I listen to the system when the speakers are wired together, and also when they’re connected wirelessly.
Wireless sources are Tidal Connect and Apple AirPlay, while a Rega Apollo CD player is connected to the digital optical input, a Philips OLED806 TV is connected via HDMI, a preamplified Technics SL-1300G turntable is connected to the analogue input, and a Synology flash drive is accessed via the USB-A slot.
This arrangement allows for access to a lot of different content, stored in a lot of different formats.
No, the size of the bottom plate of the speakers means any decent speaker stand will do just fine. It won’t have as harmonious a look, though, obviously.
No, Yamaha is convinced the system performs at its optimum when the drivers are not concealed.
| Yamaha NX-70A Review | |
|---|---|
| Manufacturer | Yamaha UK |
| Size (Dimensions) | 189 x 234 x 333 MM |
| Weight | 5.7 KG |
| Release Date | 2026 |
| Audio Resolution | Up to 32-bit/384kHz / DSD256 |
| Driver (s) | 30mm Harmonious Diaphragm dome tweeter; 130mm Harmonious Diaphragm cone mid/bass |
| Audio (Power output) | 320 W |
| Connectivity | Bluetooth 5.3; dual-band Wi-Fi |
| Colours | White, Black |
| Frequency Range | 50 35000 – Hz |
| Amplification | Class D |
| Speaker Type | Active Speaker |
| Remote Control | Yes |
| Inputs | Ethernet; digital optical; HDMI eARC; line-level analogue; USB-A |
| Outputs | Subwoofer |
| UK RRP | £2599 |
| USA RRP | $3499 |
| EU RRP | €2999 |
| CA RRP | CA$4299 |
| AUD RRP | AU$4799 |
San Francisco’s city attorney has hit Meta with a cease-and-desist letter telling the company to stop “allowing” paid advertisements that include AI-generated child sexual abuse content and asked it to explain how the ads repeatedly ran on its platforms.
On Tuesday, WIRED reported that Meta has run more than 350 ads in recent months that transformed still images of minors—some of them confirmed as real individuals, including a member of a European royal family—into short videoclips that can depict them performing sexual acts. When clicked, these ads would direct people to download AI image- and video-generation apps, some of which could be used to digitally undress people or create other nonconsensual intimate imagery.
The ads, which were discovered by researchers at the Tech Transparency Project (TTP), included some identical ads that had been uploaded multiple times, and in total they reached more than 29,000 accounts in countries in the European Union. The ads that were shown across Facebook, Instagram, and Threads also targeted accounts in the US, Australia, and India. WIRED initially reported on a batch of 53 of the ads at the start of August, but more than 250 of the total number of ads ran on Meta’s platforms after this initial report.
“Meta has failed to address a known issue of AI-generated child sexually explicit ads, while profiting from them,” San Francisco city attorney David Chiu says in a statement to WIRED. “These ads are deeply disturbing and cause harm to real children. Meta claims to have zero tolerance for child exploitation, yet the scale and persistence of these issues suggest current efforts are woefully inadequate. This is unacceptable, and Meta must do better.”
“No company—particularly one that claims all advertisements are reviewed and approved prior to distribution and then accepts payment for that distribution—should permit its advertising systems to be used in this manner,” Chiu writes in the four-page letter his office sent to Meta’s lawyers.
The letter raises a host of questions about how Meta’s ad policies and safety systems operate and asks that the company’s legal, child-safety, and advertising staff engage in “prompt discussions” with the city office. “These findings are especially troubling because they do not appear to reflect an isolated failure that Meta corrected when first alerted to the problem,” Chiu writes.
Meta says there is “no indication” that the ads were displayed in San Francisco and this means they are “outside the city attorney’s jurisdiction.” While Meta’s ad library, which is a repository of ads, includes some data about ads published in the US, it does not break down where in the US ads may have run.
In response to Meta’s jurisdictional claims, City Attorney’s Office spokesperson Alex Barrett-Shorter tells WIRED that, because “San Francisco consumers can access the unlawful ads through Meta online spaces,” the matter is the the city of San Francisco’s business.
“We find Meta’s response astonishing. Meta appears less concerned about whether they’re guilty of exploiting children than whether San Francisco should say anything about it,” Barrett-Shorter says. “Meta has bigger problems to solve here than determining our jurisdiction.”
Meta previously told WIRED that many of the ads identified by the TTP researchers had already been removed before they were reported—all of them have now been removed for violating its policies—and most had “fewer than 200 impressions” from accounts. “We don’t tolerate nudify apps or any kind of child exploitation, whether real or AI-generated,” a Meta spokesperson previously said. The company said the “total ad spend” for the 300-plus ads found by the researchers was “under $5,000.”
The cost of data centers continues to rise, with insurance now added to real estate, communications, energy, and environmental impact. Several reports indicate that not only is insurance a growing cost for big guns like OpenAI and Meta, it’s a market that looks set to expand alongside the increase in data centers.
A new report by the Swiss Re Institute indicates the insurance industry could collect $91 billion in premiums from AI data centers between now and 2030, with a further $111 billion from renewable energy installations linked to the data centers.
With a combined market of around $200 billion across a three year period, the AI boom could represent an accumulation of risk. The report highlights four interconnected factors that could affect multiple businesses if only one is disrupted.
Latest Videos FromTechRadar
Reliance on new data centers isn’t all about the end product. Construction companies, physical supply chains, and technology and communications organizations all play their part in assembling an AI data center facility. Power stations – traditional or renewable – have similar requirements, and once both types of installation are brought online, their importance increases.
Gianfranco Lot, Swiss Re’s chief underwriting officer for P&C Re, told Insurance Business Mag: “AI needs data centers, power grids, and increasingly complex infrastructure – and all of it needs insurance. That creates growth opportunities across multiple lines of business, but also significant risk concentrations. The deployment of capacity will depend on our ability to understand and manage those, and getting paid for the associated tail risk.”
Estimates suggest a single AI data center can cost $50 billion to replace.
Insurers are preparing to offer comprehensive support for the data center industry. Risk management firm Aon has released an analytics tool designed to help insurers measure and map exposures for data centers, and it appears the wider industry is exploring options to underwrite these installations.
Indeed, Aon has its own estimations, working on a figure of a $29 billion market by 2030. Meanwhile, other insurers and analysts are assembling databases and exploring opportunities to provide financial fallback to data centers.
Four factors (large assets, geographic clustering, supply-chain dependencies, and shared networks) can each interrupt several businesses within the sphere of the data centers, and the Swiss Re Institute report highlights the importance of insuring against disruption. This represents a previously unseen level of integration and interconnectivity between industries, something that seems to be informing the market estimates.
Ultimately, the insurance price will be factored into the usage costs, with AI token prices and other tariffs added to the bill for end users.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
We may not want to pay $2,000 for it, or couldn’t spend that even if we wanted to thanks to RAMaggedon-high prices, but there’s no doubt that the centerpiece of Apple’s primary iPhone launch this year was the foldable iPhone Duo. From a bread-and-butter perspective, though, the iPhone 18 Pro and Pro Max, despite a price increase, were the highlight of the Surprise and Shine event. Apple took the opportunity to raise the prices of much of the rest of the line, though.
Much of what we’re treated to during these launch events is a recap of the new capabilities of their operating systems, which have usually been available in beta since the developers conferences where they get announced.
For Apple, that means demonstrations of its updated iOS 27 and WatchOS 27, announced at WWDC in June running on the new hardware. But we’ve been using the OSes for a while and don’t have to rely on passing demos during the hardware announcement: Here’s our experience with the beta and what to know about iOS 27 before you start using it as well as our take on the new Siri in WatchOS.
On the upside, the memetic reaction to the Duo brought out the creative snarkers of social media.
Check out our complete coverage of Apple’s September 2026 iPhone launch and play-by-play commentary on the event in our archived live blog.
Preorders for the iPhone 18 Pro and Pro Max and Watches start soon and they ship a week later the Duo is up for preorder starting and ships at the end of October. And don’t forget the cases.
After the initial cinematic opening to promote the new variable aperture camera on the iPhone 18 Pro and Max, Tim Cook provided a clever, well-done segue to Ternus’ first Apple launch as CEO.
Ternus came up through product design, and it showed during the event. If memory serves, we were provided with more specific hardware information about chips, device builds, cooling systems and so on than in a lot of earlier launches. That might not appeal to everyone, but it’s the type of thing that keeps me awake as we head into the second hour of these things.
There also seemed to be a lot more video shot in lots of locations around LA. This was likely to drive home the association between Apple and “pro” cameras, but it made a very nice change from a lot of static Apple Park and heartwarming family video.
Abrar Al-Heeti/CNET
In addition to being its first foldable, the iPhone Duo brought a much longed-for feature back into the…fold.
Touch ID disappeared along with the iPhone’s Home button; the last holdout was the iPhone SE which was discontinued in 2022. Even if you’re fine with Face ID, having a fallback method for logging in and approving purchases makes sense, and its certainly more convenient than having to enter a PIN code.
But it unfortunately replaced Face ID rather than supplement it.
We had some in-chat mockery about ability to register multiple fingerprints for the Touch ID, but as someone who occasionally has dirty, greasy or otherwise electronics-objectional hands, being able to use a clean finger sounds nice to me.
I like the simultaneous front and rear camera feeds, which lets use the outer screen flipped towards the subject. It reminds me of methods introduced to get pets’ attention from the early days of phones and before phones had cameras; I remember using an app that mimicked different cat vocalizations, though I really spent a lot of time wiggling my fingers above my head.
Abrar Al-Heeti/CNET
I’m also a huge fan of the nano-texture screen layer, which comes from Apple’s desktop displays and MacBook Pros. With its excellent anti-reflective properties, it makes the most sense on a phone. I’d like to see it on more than just the Duo.
My biggest concern about the Duo was the size. The iPhone 16 Pro Max I currently have occasionally feels a hair too big to use comfortably single handed. But the folded Duo is a bit narrower.
Apple made a big deal about its mechanical aperture for the new 48MP main cameras on the iPhone 18 Pro and Pro Max, and it is a breakthrough for a phone.
Electronic apertures used by most small cameras and phones simply limit or increase the amount of light the sensor absorbs. A mechanical, bladed aperture physically changes the shape of the hole which controls the exposure, and that determines what bokeh and bright highlights look like as well as how naturally the depth of field is rendered.
So introducing a mechanical aperture should allow for more control over the quality of the look (as opposed to the quality of the image) looking photos and videos. That’s notable.
But don’t expect a truly pro-lens look, though. It’s a six-bladed diaphragm, which isn’t even as good as some entry-level kit lenses, and 8 blades is more or less the minimum for really nice-looking, soft bokeh. In its video showing off the camera, there was very little of it.
There are four aperture settings, f1.48-f4, which highlights the constraint of the small phone sensor and module. By f4 on a sensor that small you’re essential at infinite focus, where on a larger sensor you might get up to f16 or f32. The bottom line is you don’t have a lot of depth-of-field variability to play with for that variable aperture.
I’m reserving judgement on the underscreen front camera for the moment, though.
I’m a fan of deep, rich colors for phones, though I usually just cram it in a case and never see it again. At least based on the livestream, the burgundy looks like a nice change from the aggressive orange of the previous generation (though I do like it), and the navy-black Night Sky seems a lovely choice after my MacBook Pro-favorite Space Gray.
In a switch-up from previous years, Apple launched its new laptop and desktop chip, the M6, ahead of the A20 Pro, which is incorporated by all three iPhones. Usually the mobile processor comes first and the computer chip inherits its major architectural changes from that. In this case, the A20 has the same 2nm process size as the M6, down from 3nm in previous generations.
A smaller process size allows for a denser die, which ultimately lets you fit more circuitry in the same size package.
In this case, it has the doubled Neural Engine of the M6, used for the on-device AI processing, from 16 to 32 cores. That’s likely an essential upgrade for decent performance with on-device Siri AI.
There are also more cores, both GPU and CPU, plus the change from performance cores to super cores, which can run at faster clock speeds.
It also has the packaging with the memory placement optimized for better heat dissipation because smaller process sizes and denser dies can generate a lot more heat than bigger ones. That plus the switch to super cores required it.
Apple also dealt with heat by adding a vapor chamber and adding a dissipation layer beneath the screen.
Efficiencies introduced by the chip theoretically improve battery life, according to the company.
If you’ve got two iPhones, iPhone handoff will let you share a single phone number, designating one eSIM as primary and the other as a companion. It sounds interesting, and would let you leave one phone at home or, with careful scheduling let someone else pick up (for say an elderly relative or young child) but it’s one of those features that could suffer if it’s badly implemented. I’m not sure if it lets you save the cost of a second line, because you’ve still got to have an activated SIM in the companion phone. I also wish it could work with a non-Apple phone.
It’s also kind of confusing, since Apple already has Handoff for sharing tasks across devices.
More sensitive, frequently polling sensors are an unambiguous win if you use them in a watch, as long as they don’t overwhelm the rest of the system (which I have no reason to think they do here). Better sensors in the Watch Ultra 4 mean more accurate readings and more easily recognized outliers.
The Watches’ Audio Intelligence can, among other things, decode sounds into visual notifications for the hearing impaired. That’s a nice accessibility perk.
An AI-driven feature that I could really use, though, is real-time accent compensation in AirPods; the AirPods 5 received some tweaks inherited from the higher-end model, including improved active noise cancelling. Translation is nice, but I have a lot of trouble understanding people with heavy accents of all kinds — a big drawback when you live in New York City.
It’s not all sunshine, though. There are a lot of things I just say “nope” about.
On one hand, I like the idea of a Watch that can catch parts of conversations I routinely miss (Live Rewind), either because I’ve gone space cadet or because my hearing glitched, as it tends to do. But I’m also one of those Luddites who doesn’t like or use devices with always-on listening or recording.
I’m not a big fan of Apple’s “it only stays on our servers” Private Cloud Compute privacy narrative, which surfaced once again today, especially since Apple Intelligence/Siri AI can also use hybrid (local plus cloud) and third-party-integration models. Even if it stays on your device, you still never know until after the fact what has obtained access to your data — and it may now include a lot of health data.
It may also be because I don’t find a lot of the commonly trotted out use cases very compelling. Apple Intelligence, and the upcoming dedicated Siri app, can do the things everybody’s on-device AI can, which seems to be writing email, summarizing calendars, travel scheduling, creating shopping lists, generating wallpaper and so on.
To me, a lot of the AI used by the Watch Series 12/Ultra Health and related apps and on similar devices, such as Stuff, goes too far into the micromanaging-your-life zone. If it’s going to give a Readiness score of my ability to handle activities, I’m going back to bed and pulling the covers up over my head, because my body tells me it’s not ready for anything.
While the Duo has the same 48MP main camera as the iPhone 18 Pro and Max, it doesn’t have the same mechanical aperture or high-end features. This is actually a notable negative which I’d weight heavily when considering the phone.
The other thing I suspect will make things awkward are the screen aspect ratios. Folded, each side is just shy of 3:2 — it’s 1.45 instead of 1.5 — which has become nonstandard. 1.5 is the aspect ratio of full-frame camera sensors, which conceivably makes it nice for displaying them, but neither video or games use it. Opened, the large screen has a 1.4 aspect ratio, which is even more novel.
As a result, Apple has had to alter iOS 27’s screen layout settings, both to take advantage of the extra screen real estate (good), as well as to change where control appear so you can use them comfortably (bad). Apple spins these as “optimizations.” But it means inconsistent user experiences and interface conventions between the Duo and the rest of the iPhone line.
Additionally, it means that the operating system will have to adapt a lot of existing non-Apple apps and content on the fly to run right on the device. That’s always a bumpy road.
While the Duo supports the Apple Pencil, it’s just plain odd. It works with the original, USB-C Pencil rather than the more recent Bluetooth version, which seems inconvenient for carrying around and using — it requires physical pairing with the device and plugged-in charging. Plus it’s not really designed to use with the phone, since there’s no way to dock it. It seems like the least-effort solution possible to a stylus.
Bitcoin price risks $76K drop as $78K support weakens
Memory prices are slowing because buyers ran out of money
Ethereum price stalls below $2,500 as ADX drops to 11
Robinhood Stock: How To Take Advantage With Reduced Risk
Phones confiscated, players sent home: Pakistan’s England tour turmoil revives memories of Mohammad Amir, Salman Butt and Mohammad Asif’s 2010 Lord’s spot-fixing scandal | Cricket News
England up in reading, maths and science rankings as Scotland and Wales dip
Bitcoin price holds near $79K as cycle drawdowns narrow
Brent Crude Oil Moves Above $100 for the First Time in 3 Months
Meta debuts long-awaited personal AI agent, Muse
Visa expands stablecoin card network to 160 programs
Egyptian TV Presenter Sarah Khalifa, 11 Others Sentenced To Death In Major Drug Trafficking Case In Cairo
Strong Password Policy and Password Manager Guide
Intel Stock Jumps 9% on Chip Price Hike Report, US Stake Gains $36 Billion
Big Brother Spoilers: Dee SPINS & HUSTLES Votes – Yash Questions the Pact
Ballon d’Or 2026 Nominees: Lionel Messi In, Cristiano Ronaldo Out; Erling Haaland Included In Top 30
ServiceTitan, Inc. (TTAN) Q2 2027 Earnings Call Transcript
Google’s revived nuclear power plant gets $1.9B loan from US government
Meta debuts its Muse AI agent. Will consumers trust it?
Tehran’s assertive rhetoric signals a new and dangerous phase in conflict
Oil Higher on Hormuz Escalation, Weighing on Sentiment
You must be logged in to post a comment Login