News

• MGI Engineering unveils cruise missile-like drone system, capable of speeds of 466 MPH and designed to operate autonomously in GPS blocked environments
• The TigerShark can carry a 300 kg payload, with a range of 900 km
• Parent company MGI Engineering has worked with Formula 1 teams, Formula E cars, and developed a car that won at the Dakar Rally

A British engineering company usually associated with Formula 1 and other motorsports has built an affordable “one-way effector” (meaning single-use) strike drone, essentially an intermediate-range cruise missile. MGI Engineering’s TigerShark is deployed from a ground-based vehicle with a rocket-assisted launch, whereby its onboard propulsion takes over and the rocket is detached.

Capable of delivering a 300-kilogram payload, the TigerShark uses GPS for navigation but is also designed to reach its target in terrain where GPS is unavailable or jammed.

The TigerShark is intended to be launched in multiples, with a swarm of the drones intended to attack targets, aiming to overwhelm air defense systems.

Based in Witney in Oxfordshire, MGI Defence — part of MGI Engineering — first unveiled the TigerShark at the DSEI exhibition in 2025. It now aims to bring the TigerShark to market within months, aiming to sell it to Britain’s strategic allies as a drone-based alternative to traditional cruise missiles.

Designed for faster assembly and delivery, the TigerShark can solve the problem of shrinking defense budgets, too, costing far less than Storm Shadow or Tomahawk cruise missiles.

As a one-way effector, the single-use drone is designed with a rocket-assisted propulsion for launch while twin turbojet engines drive the 661-pound payload to its target up to 900 kilometers (559 miles) away. Reaching speeds of up to 750 km/h (466 mph), the TigerShark can operate despite GPS jamming, thanks to inertial navigation.

This is one aspect of a pair of technologies used in conjunction to assure GPS-free targeting. While inertial navigation tracks position based on motion and speed from the launch point, a terrain-mapping system ensures what the TigerShark sees below matches its intended route, and makes adjustments. So, even if GPS is jammed, the TigerShark should reach its target.

MGI has announced that the drone is available from October 2026, and will be manufactured in the UK with export licenses to friendly countries.

The company states that “TigerShark enables defence forces to engage defended or distant targets with speed, precision, and autonomy, without relying on satellite navigation or real-time control. Its guidance suite combines inertial navigation with GPS-free terrain mapping for reliable operation in GPS-denied and spoofed environments.”

“Built for coordinated launch and rapid deployment, TigerShark can be fired as a salvo from a land-based vehicle launcher to overwhelm or degrade enemy defences. Its modular payload bay supports a wide range of mission types, from warhead delivery to electronic disruption.”

The TigerShark will enter increased production in late 2026, with MGI Defence aiming to sell to allies needing a more affordable and faster acquisition alternative to traditional intermediate-range cruise missiles.

• The Cherokee Nation has placed a moratorium on new data centers on tribal lands
• Citizens say they are concerned about the impact of new projects on the local areas
• Data centers projects on non-tribal land will not be granted support by the Cherokee Nation unless they are given early consultation

Hyperscale data center development has been banned on tribal lands owned by the Cherokee Nation following a poll of its citizens.

The poll was run by a task force headed by Principal Chief Chuck Hoskin Jr., and found that 64% of the 1,593 respondents were opposed to the construction of data center campuses within Cherokee Nations lands, which cover about 6,963 square miles across 14 counties in Oklahoma.

Citizens' concerns returned in a 37-page report focused heavily on water consumption, air quality, noise, and cultural resource protection.

The report also stated that data centers seeking approval to be built on non-tribal land within the reservation will require an early consultation with the Cherokee Nation, and will not be supported unless this early consultation takes place.

"Our primary responsibility is to protect our citizens and tribal communities from these threats, so we will not support any hyperscale data centers on our reservation without proper consultation," Hoskin said.

As there are no Oklahoma state rules requiring data centers to register, the number of planned and operational data centers is unknown. But estimates vary. CleanView places the number of planned and operational sites at 41, while the US Data Center Map says there could be as many as 51.

What we do know is that every hyperscale project within the state is on non-tribal land. Google’s Pryor campus has operated since 2011, but is undergoing expansion as part of a $2 billion investment in the Stillwater and Pryor campuses.

The citizens of the Cherokee Nation seem to share the same concerns as many other Americans, with fears of job replacement, environmental damage, and rising energy bills also being prevalent as reasons for AI data center opposition.

Other tribes, such as the Seminole Nation and Kickapoo Tribe, have already passed their own bans on data center projects within their territory.

Via Tom's Hardware

• Ordinary Wi-Fi networks can reportedly identify people without cameras or consent
• KASTEL researchers achieved nearly 100% identification accuracy across 197 volunteers tested
• Wi-Fi signals can reportedly reveal people from different viewing angles

A new study suggests that ordinary Wi-Fi networks could be repurposed to identify individuals without cameras, sensors, or their knowledge or consent.

Researchers from KASTEL, the Institute of Information Security and Dependability at KIT, tested the approach on 197 volunteers in controlled conditions.

Their system reportedly achieved nearly 100% identification accuracy regardless of a person's viewing angle or the way that person walked.

The technique relies on beamforming feedback information, a type of signal that connected devices routinely send back to a router.

This information is transmitted without any encryption, meaning anyone within range of the network could potentially intercept and read it.

By analyzing these radio signals over time, the system reportedly builds images of people from multiple viewpoints inside a space.

Earlier wireless sensing methods often required specialized equipment, such as LIDAR sensors or detailed measurements of channel state information from a network.

By contrast, the researchers say a standard, unmodified Wi-Fi device is entirely sufficient to carry out this new identification process.

"By observing the propagation of radio waves, we can create an image of the surroundings and of persons who are present," said Thorsten Strufe, a professor at KASTEL, KIT's Institute of Information Security and Dependability.

"This works similar to a normal camera, the difference being that in our case, radio waves instead of light waves are used for the recognition.

A user does not need to carry a phone, smartwatch, or other Wi-Fi-enabled device for the system to work.

Turning off one's own device would reportedly not prevent identification, since other active Wi-Fi devices nearby could still be used.

The researchers argue that widespread wireless networks could create privacy concerns because they operate throughout homes, offices, restaurants, cafés, and other public environments.

Felix Morsbach noted that intelligence agencies and cybercriminals already have simpler methods, including compromising CCTV systems and connected video doorbells.

“The omnipresent wireless networks might become a nearly comprehensive surveillance infrastructure with one concerning property: they are invisible and raise no suspicion,” Morsbach added.

Compared with existing surveillance options, such as CCTV footage or connected video doorbells, wireless signals leave no visible trace behind.

Professor Strufe cautioned that this same invisibility could make the technology particularly attractive within certain authoritarian political systems around the world.

"The technology is powerful, but at the same time entails risks to our fundamental rights, especially to privacy," said Strufe.

Ordinary routers could potentially enable people to be recognized without their knowledge during routine movements, and this could be dangerous in the wrong hands.

Therefore, widespread deployment could make individual encounters difficult to distinguish from ordinary wireless activity.

The researchers argue that privacy safeguards should be incorporated into Wi-Fi technology before such capabilities become widely and easily exploited.

Specifically, they are calling for protections to be written into the forthcoming IEEE 802.11bf wireless standard currently under active development.

Via ScienceDaily

• Researchers find a standardized SIM command is exposed on nine of 26 tested devices and used it to achieve code execution on a commercial EV charger
• The exposure is concentrated in machine-to-machine hardware rather than phones, affecting six of eight cellular modules but only three of 18 handsets, with no iPhone or Pixel among them
• Every attack requires the attacker to already control the SIM, and while Qualcomm has produced a hardened configuration disabling the interface by default, no vendor had published a public advisory yet

A malicious SIM card can instruct the device it sits in to run commands of an attacker's choosing, and on the cellular modules embedded in electric vehicle chargers, industrial routers, and car telematics units, essentially allowing it to take the entire device over.

Researchers from the University of Birmingham and the German security firm Fuzzware demonstrated this against a commercial Autel EV charger, achieving code execution driven entirely SIM card-issued commands.

The work focuses on a standardized feature called Proactive SIM, which as a feature, is not malicious; it's a standard in a cellular specification that lets a SIM push commands to a device rather than acting as a passive identifier for one's identity on a network.

The problem is one specific command in that set, RUN AT, which asks the modem to execute an AT command, the modem control language dating to the 1981 Hayes Smartmodem that every vendor has since extended with its own additions.

The support extender essentially gives a SIM module its own general-purpose console on devices that lack safeguards to prevent such an attack.

Tomasz Piotr Lisowski and Dr Marius Muench of Birmingham, working with Fuzzware's Kristian Covic, built a toolkit called CATana to find out what a hostile card could do with that console.

The team tested 26 devices, 18 smartphones and eight cellular modules, and found the SIM AT interface exposed on nine of them. The exposure is overwhelmingly concentrated in machine-to-machine hardware: six of the eight modules accepted the command, compared with just three of the 18 phones: the Oppo Find X5, the Oppo Reno 14 F 5G, and the Asus Zenfone 9. This makes it not exactly a Simjacker-esque exploit but still one that needs to be taken seriously.

All nine devices that accepted the command run a Qualcomm chip or modem, but five others that do were not vulnerable to the attack. The researchers first shared the reports with Google, Oppo, Quectel, Semtech, and Qualcomm in March 2026, and with the GSMA in May. Qualcomm has since built a hardened configuration that switches the interface off by default, which the researchers say will be the default on future devices.

The attack vector, however, is limited because, to leverage it, the attacker must already control the SIM itself. Knowing a victim's contact number is not enough; the attacker needs physical access to the SIM slot. The exploit is real and concerning, but it has limited utility compared to a remote one for smartphones.

The IoT side is more concerning, however: unattended equipment with an accessible SIM tray can now potentially be exploited, and physical swaps might be easier than with something more personal, like one's personal phone. The card essentially talks to the equivalent of a Linux computer, one that the paper calls a rich attack surface for hostile SIM cards.

For now, the irony is that while modern smartphones have largely retired the surface this attack vector exploits, the machine-to-machine world has not, and the equipment least likely to receive a firmware update is the equipment most exposed currently.

Many pioneering pieces of fiction in the 21st century have come to accurately define the relationship between humans and AI before the contemporary explosion of modern intelligent systems. These include the film 'Ex-Machina', various episodes of 'Black Mirror' and, of course, 'Her'.

Starring Joaquin Phoenix and Scarlett Johansson as the AI agent Samantha, 'Her' was a fresh and mature approach to the way that humans and AI may coexist in the future.

This article is part of TechRadar Pro's QOTD project to provide an insight into the minds of the brightest and most recognized figures in the technology industry today and in years gone by. Read the full series here.

Ten years after its release, with ChatGPT gaining popularity and transformer-based AI improving on every level, Altman approached Johansson's agent to ask permission to use her voice for the marketing of GPT-4 on its planned launch in 2024.

Johansson declined, but Altman pushed on with a voice, dubbed 'Sky', that was near-indistinguishable from the actor's, essentially suggesting the technology that OpenAI had created was on a par with the fictional tech in the film. That's why, in a statement, Johansson expressed her anger and intention to get the voice taken down.

Johansson considered the offer, with Altman telling her that her voice "could bridge the gap between tech companies and creatives". It would also "help consumers to feel comfortable with the seismic shift concerning humans and AI".

She hired legal counsel, and OpenAI agreed, after two letters, to take down the 'Sky' voice. The incident raised deep concerns over the early potential for AI to replace the likeness of actors, which could undermine their careers and livelihoods.

It was also a dark reminder that deepfakes are becoming increasingly easy to create, and that it's not only harmful users, but AI companies and executives that can so brazenly co-opt another person's identity for their own purposes.

• Ukrainian civilians and bereaved families sued Ubiquiti in Manhattan federal court
• Lawsuit alleges the communications backbone offered by Ubiquiti's long-range wireless bridges was a key part of Russia's drone campaign against civilians
• Ubiquiti has not responded directly or commented on the matter to date and remains uncharged by any US or EU authority at the moment

Ukrainian civilians and relatives of people killed in Russian drone attacks have sued American networking manufacturer Ubiquiti Inc. in the US District Court for the Southern District of New York.

The lawsuit alleges that the company's long-range wireless equipment became the backbone of the command network behind Russia's drone campaign against civilians in occupied and front-line Ukraine.

Ubiquiti has not filed a response or commented publicly; it has also not been charged by any US or EU authority with a sanctions or export-control violation related to Ukraine at the time of writing.

The Russia-Ukraine war has lasted over four years, and it has reshaped the modern battlefield through asymmetric approaches from both sides. Militaries worldwide are watching as they learn what modern warfare could look like in the near future between two neighboring countries.

Drones, however, are a key part of the battlefield, and while both sides employ jammers and countermeasures to said jammers, including fiber-optic drones, they remain among the most destructive devices in the conflict.

Wireless drones, however, still rule the roost, especially for deeper strikes behind enemy lines, which is why an NYSE-listed company (ticker UI) now finds itself mired in an unexpected lawsuit over its radio bridge antennas, which are also subject to US export restrictions on Russia.

Ubiquiti makes long-range wireless bridges: paired antennas that carry a network link between two distant points, widely used to deliver internet access where cable does not reach.

The complainants allege that devices circumventing export controls managed to keep reaching Russian forces through intermediary supply chains, and that they now link reconnaissance units, drone operators, strike units, and commanders across roughly 100 kilometers of battlefield network.

They also alleged Ubiquiti could have cut the pipeline off through geofencing, by terminating distributors tied to sanctions-evasion routes, or by building a remote-disable feature into its firmware, and that it chose not to take any of these initiatives.

Scott Gilmore, partner at DiCello Levitt and lead counsel for the plaintiffs, said in the firm's announcement of the filing, "This case is about corporate accountability for enabling crimes against humanity."

This is despite Ubiquiti conceding that it doesn't have "any visibility on the location or extent of purchases" made by individual network operators and service providers from its own distributors.

The approach that the lawsuit hints at is not unlike how SpaceX handled its Starlink terminals once Russian forces started using them after it initially denied they were being used systematically. Government-level intervention eventually forced SpaceX to require users to register their units on an official government-approved whitelist to restore connectivity.

If the motion passes, it would considerably change how supply-chain accountability litigation is handled in the future for a company that did not directly sell them to a sanctioned entity, especially as the White House is in an uproar over bleeding-edge Nvidia AI chips making it to China despite export restrictions. Ubiquiti, on the other hand, could argue that it has no duty to police where its products are re-exported and that no duty runs from a networking manufacturer to civilians in a war zone thousands of miles from the point of sale.

• Alibaba Cloud's Lingjun Zhenwu M890 supernode instance is now on commercial sale
• It offers rentable 64-card computing units built entirely on in-house silicon
• It marks the first time Alibaba has directly sold supernode compute through the public cloud, however availability is limited to the Ulanqab region in Inner Mongolia

Alibaba Cloud has begun commercial sales of its Lingjun Zhenwu M890 supernode instance, letting customers rent a 64-card, high-speed interconnected computing unit without building their own data center.

The launch makes it the only Chinese-origin supernode that users can effectively use online without entering a purchase agreement or using a preconfigured AI model instance, an impressive enterprise-centric addition.

There is a caveat for enthusiasts looking to test it outside China: it is currently limited to users in North China's Inner Mongolia Autonomous Region only.

A supernode is a set of accelerators wired together tightly enough that software treats them as one very large chip rather than a network of separate ones. This approach is often necessary; a model too big to fit on a single card would otherwise have to shovel data across comparatively slow networking, and that overhead can determine whether trillion-parameter models are practical to serve.

One of the best examples is Nvidia's GB200 NVL72, which leverages 72 of its Blackwell GPUs in a single configuration, or AMD's 72-GPU Helios configuration, which offers its own Instinct MI455X GPUs with more memory per GPU in tow.

Alibaba's offering is not faster than either example, and both are among the fastest options available to consumers worldwide, but it offers a mix of advantages to domestic Chinese consumers that would not otherwise be in play: Chinese Silicon, a Chinese cloud, and lower latency.

As Beijing restricts Chinese cloud providers from accessing the latest Nvidia and AMD chips they can buy in a tit-for-tat move for the White House's increasing heavy-handedness, the best options for enterprise customers and AI startups have become increasingly domestic, and Alibaba Cloud's homegrown silicon is a key example of how that policy has shaped its offerings.

The underlying Zhenwu M890 processor debuted at the Alibaba Cloud Summit in May 2026, where the company also disclosed a detailed chip roadmap through 2028. comes from T-Head, the group's chip design unit, and carries 144GB of HBM memory with inter-chip bandwidth up to 800 GB/s. The interconnect fabric uses an in-house ICNSwitch 1.0 chip, which lifts the scale-up domain from 16 cards on the previous-generation Zhenwu 810E to 64, with bandwidth up from 700 GB/s.

The company says that the configuration is, on paper, capable of handling a Mixture of Experts (MoE) AI configuration of up to 10 trillion parameters, while emphasizing that this is the first supernode-form architecture in China to successfully run large language models exceeding 2 trillion parameters; Alibaba says the current instance is already powering commercial services for Kimi K3 and Qwen3.8-Max.

Alibaba's roadmap continues past the M890. T-Head has said the Zhenwu V900 arrives in the third quarter of 2027 with three times the M890's performance, 216GB of memory, and 1,200 GB/s of inter-chip bandwidth, followed by the Zhenwu J900 in the third quarter of 2028.

There is no public pricing for the instance, and information on its power consumption is nonexistent; benchmarks comparing it to mainstream options are scant at best. Despite the limited information and geographic limits, the move underscores both China's growing self-reliance in the AI industry and its ambitions, which could see the supernode offered to a wider range of consumers, both domestic and international, in the near future.

Jack Reacher is the action hero that I never expected to fall in love with — but if you've been a long-time fan, you probably knew that would be the case long before I did.

As of writing, Reacher season 4 is dominating the Prime Video charts, and for good reason. Without giving anything away, the action has ramped up in a way that the Amazon TV adaptation has never seen before, and with a fifth season having already been renewed, things aren't about to slow down any time soon.

But things didn't start with a TV show a few years ago, did they? The character of Jack Reacher has been around for much longer, and has quite literally taken on a variety of different faces.

Here's where the challenge comes in. Whether you're a Reacher newbie or seasoned veteran (military pun intentional), just how much do you know about the man behind the action legend? Take the quiz below to find out:

As a side note, you'll need to have seen Reacher seasons 1-3 to get as many questions as possible correct, alongside having at least a baseline understanding of season 4.

While there's some crossover with the book series, questions are primarily based on the Prime Video series — I am a streaming writer, after all.

Even so, I haven't gone easy on you... I'd bet any imaginary money that you won't score about 70%. Prove me wrong, dear reader, prove me wrong.

• Google is building on its new Insulin Resistance Trends health feature with PhotoScan
• PhotoScan is 'an investigational deep learning framework' that can estimate body fat using a phone's camera with 'near DXA accuracy'
• It's not yet available to the public, but the research 'demonstrate[s] the feasibility' of such features

As the Google Pixel Watch 5 launched last week alongside a slew of new Pixel phones, one of the most interesting new developments was the introduction of the Pixel Watch's new Health Guardian suite of tools, including Insulin Resistance Trends.

Insulin Resistance Trends showcases a month-by-month breakdown containing estimations of how your body responded to blood sugar spikes — a crucial component in weight management and the development of Type-2 diabetes.

Body composition — your ratio of different kinds of fat, skeletal muscle, and bone — is also an excellent marker of insulin resistance and metabolic health, so body scans are often used to help determine insulin resistance. Body comp is often measured with medical-grade 'DXA' scans, which use X-Rays, or commercially available devices at home.

These home devices can be the best smart scales or even some smartwatches, such as the Samsung Galaxy Watch Ultra 2. Without medical DXA scanners, they usually use a process called 'bioelectrical impedence analysis', or BIA. A low-level electric current is sent through the user's body, moving through fat, muscle, and bone at different speeds. The speed of the electrical current helps the device estimate the percentages of each material in the body: a useful thing to have on a health-monitoring device like a smartwatch.

However, Google is busy researching an even more accurate version of a body scan without using any specialist equipment or BIA kit — in fact, it's just a smartphone camera and an onboard AI model. This in-development feature is called PhotoScan, appropriately.

According to its research blog, Google has been developing PhotoScan by matching smartphone photographs of subjects with medical-grade DXA scans, along with other subject information, then running it all through an AI model. The AI model eventually achieved 'strong DXA agreement' when estimating body fat percentage and other metrics, leading to 'near-DXA accuracy for predicting insulin resistance'.

The researchers conclude that 'clinical DXA imaging delivers the most accurate body composition but lacks scalability, whereas wearable BIA sensors offer convenience but are limited to basic body fat percentage.

'Our PhotoScan approach offers a promising middle ground, estimating granular body composition from standard smartphone imagery with near-DXA accuracy.' Sorry, Samsung — it looks like we'll be pivoting away from BIA sensors and towards smartphone cameras for body fat percentage calculation in future.

While the technology doesn't seem to be commercially available yet, it does 'demonstrate the feasibility of smartphone-based body composition estimation,' and I'd be surprised not to see it make its way to the best Pixel phones eventually.