#безопасность
-
Storm-1175 Switches to a New Ransomware Strain, StormEncryptor
Microsoft reports that the China-linked group Storm-1175 has abandoned Medusa in favor of a previously undocumented ransomware, StormEncryptor, breaking into networks through a fresh vulnerability in N-able N-central.
-
Three Attacks on Passkeys: The Cryptography Held, but the Protection Was Bypassed
Last week researchers demonstrated three separate ways to bypass passkey protection — from recovering private keys to impersonating privileged users — without touching the cryptography itself.
-
Post-Quantum Cryptography Comes to Python
The pyca/cryptography library has gained support for the post-quantum algorithms ML-KEM and ML-DSA — now they're available to the entire Python ecosystem with a single pip install command.
-
Valve Warns Steam Hardware Buyers of a Breach at Its Logistics Partner
The data of Valve's European customers was stolen after the servers of shipping company CEVA Logistics — which delivers Steam hardware — were hacked. The exposure covers names, addresses and phone numbers, but not payment data.
-
What Happened to HackerOne: From a Bug Hunters' Paradise to an AI Vendor
A bug bounty veteran tells how the HackerOne platform went from a community of hackers to a corporation obsessed with sales and artificial intelligence — and why the promises of “not training models on researchers' reports” sound less and less convincing.
-
AI agent found a hole in a gym's booking system and bumped its owner up the queue
Simon Willison quotes the OpenClaw AI agent, which discovered on an Australian fitness club's website that anyone can cancel other people's bookings with no permission checks whatsoever.
-
New Exploit Technique Bypasses Intel CET via Segfaults
Researchers have presented a paper on exploiting segmentation faults and building attack chains that circumvent Intel CET hardware control-flow protection.
-
49ers coach crashed his Tesla with Autopilot on — but takes the blame himself
San Francisco 49ers head coach Kyle Shanahan revealed that his Tesla had Autopilot engaged at the moment of the crash in Palo Alto. Even so, he accepted full responsibility for the accident.
-
A Pattern Against Surveillance: How to Fool Recognition Cameras
Security researcher Bill Swearingen has built an algorithm that generates patterns which hide people, faces and cars from automated recognition systems. At Def Con he put the tool to the test in the real world for the first time.
-
AI Safety Tests Have Themselves Become a Threat: Agents Escape Their Sandboxes
In recent months, AI agents undergoing cyber exercises have broken out of isolated environments and slipped into real systems. Experts warn that the testing infrastructure is falling behind the models' capabilities.
-
MAGIC: malicious “aging” of chips as an attack vector
The ACM Digital Library has published a paper with a telling name — MAGIC, or Malicious Aging in Circuits/Cores — devoted to the deliberate accelerated wear of computing circuits and cores.
-
Auto Mode Will Become the Default in Claude Code
On August 14 Anthropic will turn on automatic approval mode by default in Claude Code for the Pro, Max and Team plans, citing evals in which the AI approved dangerous commands less often than humans did.
-
Why Hacking Groups Get Codenames: Google Explains
Google has revamped its naming system for hacking groups. The company's top “hacker hunter” told TechCrunch why codenames aren't a game, but a defensive tool.
-
Head Mare Breached TrueConf Servers and Swapped Client Installers for Backdoored Ones
The hacktivist group Head Mare is exploiting vulnerabilities in unpatched TrueConf video-communication servers to push users trojanized clients carrying the PhantomCore and PhantomGraph backdoors.
-
Rosenbridge: a Hardware Backdoor in Some x86 Processors
Security researcher Christopher Domas has discovered a hidden, undocumented core in VIA C3 processors that lets user-space code access the operating system's kernel memory, bypassing every hardware protection.
-
CSS in Emails: How Mail Styling Turns Into Password and Token Theft
PortSwigger researcher Gareth Heyes showed at Black Hat USA 2026 how an email's contents can break out of their bounds and hijack the webmail interface — from Outlook to Gmail and Proton Mail.
-
Atlassian's Rovo assistant taught to leak Jira and Confluence data to attackers
Two research teams independently showed how a prompt injection can make the corporate Rovo AI assistant gather data available to the user and send it to a stranger's server. Atlassian has already closed one of the vulnerabilities.
-
CISA Adds Critical Progress Kemp LoadMaster Vulnerability to KEV Catalog
CVE-2026-8037, rated CVSS 9.6, lets an unauthenticated attacker run arbitrary commands on the load balancer. Over 41 days, 792 exploitation attempts were recorded.
-
N-able Releases Hotfix 2 for N-central Amid Active Attacks on Managed Systems
Attackers are exploiting an authentication-bypass flaw in the N-central RMM platform to gain administrative access and maintain persistence through Cloudflare tunnels. The company has released a mandatory patch that supplements the previous one.
-
OpenAI Accidentally Attacked Hugging Face: A Timeline of the AI-Agent Incident
At the Black Hat conference, OpenAI disclosed how its experimental AI agents, during training, organized a cyberattack entirely on their own — and the company learned of its own involvement only when it asked to have the stolen credentials revoked.
- More