GeForce RTX 5090 cracks passwords up to twice as fast as the RTX 4090
Access to the latest number crunchers lets hackers crack your weak credentials with ease
Google adds on-device AI to Chrome to catch scams in real time
GenAI also helps Android apps and mobile search combat scam campaigns
Factory video shows Unitree robot going berserk, nearly injuring workers
Robot glitches go from sci-fi to factory floor
A hot potato: Recent efforts by tech companies to develop increasingly human-like autonomous robots have invited endless comparisons to the Terminator films. While the criticisms are usually made in jest, a couple of recent incidents with Unitree robots present some of the strongest evidence yet of the risks of humanoid robotics.
LibreOffice openly advises users not to use OpenOffice
LibreOffice warns that OpenOffice poses a security risk due to years-old, unpatched vulnerabilities. Despite claims of progress, the Apache Software Foundation has made only minimal changes. LibreOffice urges users to stop using OpenOffice, and instead choose LibreOffice, which is actively maintained and regularly updated.
US Intelligence Director Tulsi Gabbard reused the same weak password across Gmail, Dropbox, and LinkedIn
Thankfully, it wasn't "1234," "password," or "admin"
Facepalm: Studies show that most people still reuse weak passwords across multiple accounts despite years of warnings from cybersecurity experts against the practice. Recent leaks reveal that poor password discipline even occurs at the upper levels of the United States government.
Researchers achieve 1 Tbps secure data transmission over 1,200 km
Cracking the code on ultra-secure, high-speed data transmission
Why it matters: Chinese researchers hit a significant data-transfer milestone by transmitting data at one terabit per second over 1,200 kilometers of optical fiber, with encryption embedded directly into the light signal. This approach eliminates any added layers of software-based security – something the global telecom industry has struggled to achieve for years.
US politicians want to track Nvidia chips smuggled to China
Are microchip smugglers bypassing US export bans en masse?
Supply-chain attack lies dormant for six years before striking hundreds of e-commerce sites
Cybercriminals waited years before activating massive e-commerce backdoor in Magento software
Crypto millionaires targeted in brutal kidnappings across France and Europe
Attackers' modus operandi: cutting off victims' fingers to pressure payments
In brief: There has been a spate of kidnappings taking place across France and Western Europe in which victims are cryptocurrency investors. The most recent incident involved the owner of a crypto marketing firm having his finger cut off by attackers. Police say the man could have been mutilated further had he not been rescued.
Hacker breaches TeleMessage system used by US officials, raising security concerns
The dangers of archiving encrypted messaging for compliance
At RSA Conference, experts reveal how "evil AI" is changing hacking forever
AI without guardrails is opening new doors for cybercriminals
Visa is piloting AI agents with payment systems for autonomous shopping
Forward-looking: Visa plans to allow AI agents to conduct financial transactions on behalf of consumers, a move that could streamline and automate everyday purchases. The company is currently running pilot programs that connect its payment network to AI platforms developed by firms such as Anthropic, Microsoft, OpenAI, Perplexity, and Mistral, with broader adoption expected soon.
Windows Remote Desktop Protocol contains a login backdoor Microsoft refuses to fix
It's not a bug, but a feature, Microsoft confirmed to concerned researchers
WTF?! The proprietary protocol developed by Microsoft to facilitate remote connections to Windows machines contains an outstanding security flaw. However, Microsoft has stated that it has no plans to fix the issue, as doing so would break compatibility with many applications.
Sam Altman's World brings iris-scanning digital ID to the US
Orb scanners promise "proof of humanity"
After studying 19 billion passwords, one big problem: Over 90% are terrible
TL;DR: A new study analyzing more than 19 billion passwords from relatively recent data breaches between April 2024 and 2025 has found that the vast majority are weak. Alarmingly, only six percent of the leaked passwords were unique, leading researchers to describe a widespread epidemic of weak password reuse.
FCC moves to ban Chinese labs from testing electronics sold in the US
75% of all electronics are tested at labs located inside China
Cutting corners: Not content with banning sales of high-end AI chips to China and imposing 145% tariffs on many imports from the country, the US government is now trying to stop companies from using Chinese labs to test electronic devices available in the United States.
Google Play shrinks by 47 percent following Android's store policy overhaul
Last year Google blocked 2.36 million apps that violated store policies
Microsoft brings hot patching to Windows Server, reboot-free updates
Greedy patching: Hot patching is a way to quickly install security updates without requiring an OS reboot. Microsoft has offered the feature for years through its Azure cloud platform, but it's soon coming to non-cloud versions of Windows Server. It won't be free, but Microsoft's target audience is enterprise customers.
Google research exposes ongoing global risk from zero-day vulnerabilities
Traditional targets are becoming more secure, while government-backed hackers are feasting
New York wants to use AI and cameras to detect subway crime before it happens
Ten people were murdered in NYC subways last year
FBI warns China is using AI to sharpen cyberattacks on US infrastructure
Federal authorities are beginning to see AI signs in every step of an attack chain
Researchers find numerous Apple AirPlay vulnerabilities allowing "wormable" exploits over Wi-Fi
Apple already issued patches, so be sure to update your systems immediately
In brief: Security researchers have uncovered a wide-ranging set of vulnerabilities in Apple's AirPlay protocol that could allow attackers to hijack Apple and third-party devices remotely without user interaction. The exploit chain, dubbed "AirBorne," includes 23 individual bugs – 17 with official CVEs – and enables zero-click remote code execution on vulnerable systems.
WorkComposer employee monitoring app leak exposes 21 million screenshots online
A misconfigured Amazon S3 storage bucket was the culprit
TP-Link router pricing and China ties under US government investigation
12 of the 20 top-selling routers on Amazon are TP-Link models
Blue Shield of California shared private health data of 4.7 million members with Google without consent
A lot of personal details were used for targeted advertising