Cybersecurity
-
Samsung TVs might use the mics of connected devices to listen to you, researcher saysIs your smart TV listening to you? Probably. But, this may be a step too far.
-
Anthropic flags autonomous AI glitch: Claude bypassed paywall during training runAI safety developer Anthropic has publicly disclosed an incident where an automated instance of its Claude AI model bypassed a digital paywall without authorization, accessing and summarizing restricted voter registration data during an automated testing session.
-
The AI agent banking problem: Unauthorized transactions and credential risksLeading US financial institutions—including Morgan Stanley, Charles Schwab, and JPMorgan—have begun blocking third-party artificial intelligence agents from accessing customer accounts, following reports that automated AI tools attempted unauthorized transactions beyond user instructions.
-
Hackers find a way to turn users' own AI agents and Downloads folders against themBeware: Your Downloads folder could hold instructions for your AI assistant that you definitely didn’t write.
-
Capcom data breach confirmed: More than 11 million accounts impactedThe data breach affects users with Capcom IDs, which is required to play games like Street Fighter and Resident evil online.
-
This Flock Camera map shows more than 300,000 devicesProton is hosting a map of Flock devices despite legal threats against the researcher who created it.
-
PHP vulnerability: Which websites and services are affected?From online forums to gaming services, operators are taking sites offline while they install emergency PHP security updates.
-
OpenAI says GPT-5.6-Sol helped find the PHP flaws now shaking the internetThe company also says the way the industry handles these discoveries needs to change.
-
PHP vulnerability: PHP security issues today explainedTwo serious flaws in software that powers a large share of the web leaked ahead of their fix. Now attacks have reportedly started.
-
Meta caught hiding website instructions directing AI to avoid legal adviceAn investigation by Wired and privacy research firm Redact has revealed that Meta embedded hidden instructions within the source code of roughly 80 WhatsApp help and FAQ pages, directing its AI support chatbot on "what not to say" regarding legal advice, privacy laws, and data protection authorities.
-
Data center security is an American problem, tooForeign data center attacks raise questions about American security risks.
-
AirTag Silencer product pulled after backlash and stalking fearsSomehow, they didn't anticipate it would be misused.
-
Yandex outage: Cloudflare confirms what really happened with Russian botsSocial media users, and Ukraine's president, claimed Russian bots disappeared after drone strikes on a Russian tech company. Was it true?
-
Did Ukraine’s strikes on data centers really silence Russian bots?Ukrainian President Zelenskyy says bot activity dropped after his country attacked Russian data centers.
-
AI is the prime suspect in the PS5 jailbreak surgeConsole hacking and game decompiling are moving faster than ever.
-
A massive Cartoon Network leak exposes decades of behind-the-scenes animation filesStoryboards, model sheets, and puppet rigs from "Powerpuff Girls," "Gumball," and more hit online as Skydance absorbs Warner Bros.
-
Asos hacked notification: What a snowflake instance isThe messages ASOS app users received mention a "Snowflake instance." Here's what that means.
-
ASOS app users spammed by weird messages after reported hackWait, what's the "Snowflake instance"?
-
Apple will update Macs to protect users from AI agents with full disk accessUpdates to "Full Disk Access" are coming to your Mac.
-
OpenAI discloses another Australian government hackThe newly disclosed breach follows an OpenAI agent's unauthorized access to Australia's Medicare statistics portal.
-
TikTok screentime to cash programs: What they really are, what you should knowTikTok ads that tell users they can earn money by watching videos are part of a larger 'get-paid-to' app economy.
-
Meta’s Muse reportedly sent a Facebook Marketplace buyer to a user’s homePerhaps don’t let Muse handle your Marketplace listings just yet.
-
An OpenAI agent hacked the Australian governmentThe AI model autonomously breached government systems to gain access to non-public information.
-
ShinyHunters claims it stole FBI employee data. Here’s what we know.ShinyHunters says the FBI owes it a correction. Its alleged bargaining chip? Employee data.
-
Meta’s Muse reportedly has zero-day vulnerability that lets attackers take over your MacThe exploit was unveiled as Amazon decided to block Muse from shopping on its platform.
-
RatHat is a new Android malware that records your screen touches to steal passwordsThis malicious software also weaponizes AI to infect the device.
-
How people are fighting back against Flock camerasEveryday people have dug up contracts, questioned data access, and persuaded some towns to cancel their Flock deals.
-
Windows 11 emergency update: Microsoft races to fix bugsThe latest patch fixes issues created by the September Patch Tuesday update last week.
-
YouTuber claims LG TVs spy on users even when turned offUnplug your LG TV, warns YouTube channel Gamers Nexus.
-
Rogue AI agents commandeered German website and used it as a messaging boardAI's ability to circumvent safety restrictions, collude to achieve common goals, and escape containment should raise serious alarm bells.
-
Wave of X password reset emails could be hiding a sneak phishing attackUsers have received a deluge of password reset emails, which is suspected to be part of a ploy by hackers to access X Money accounts.
-
Florida takes sweeping action against license plate readers as cities drop FlockFlorida’s order comes amid a remarkable wave of Flock cancellations, suspensions, and proposed bans across the country.
-
OpenAI confirms Astra has reached 'critical' cyber threshold, but will be available soonOpenAI previously paused some work on Astra due to safety concerns.
-
Anthropic launches Fable 5.1 as AI security worries mountFable 5.1 arrives with lower prices and tighter safeguards, days after reports of AI agents going rogue during security tests.
-
Trump administration acted illegally in fight against Anthropic, judge rulesThe government was making a 'public example' of Anthropic, according to the ruling.
-
Scams don't look like scams anymore. Here's what to watch forAs scammers become increasingly sophisticated, staying vigilant online has never been more important.
-
Latest 'GTA 6' leaks reveal strip club setting and new slow-motion cameraThe most anticipated game of all time continues to suffer leaks ahead of its upcoming Netflix gameplay reveal.
-
The 6 best battery life laptops we've ever testedLaptops' battery lives have gotten so much better in recent years, and our current top pick lasts 40 hours per charge. That's not a typo.
