The day's cybersecurity headlines from public sources, in one place. Search them, filter by theme or CVE, or show only vulnerabilities CISA has confirmed are being exploited. Every headline links to the original.
For my own read on what the news means for your business, see Insights.
-
ToxicPanda Android banking trojan abuses VPN permissions to block Google Play
ToxicPanda has expanded to target 349 applications and now supports 167 remote commands. The notable change is its use of Android VPN permissions to cut the device off from Google Play, blocking Play Protect updates and remediation while the malware operates. Related Zimperium research on…
-
Encrypted prompts bypass AI safety guardrails in Grok and Gemini
Adversa AI describes a technique it calls Cryptographic Context Injection, which hides malicious instructions until they are decrypted inside a trusted execution environment, past the point where guardrails inspect input. SecurityWeek's coverage extends the finding to Gemini alongside the Grok…
-
Lawmakers ask GAO to examine the impact of CISA workforce cuts
Five House Democrats, led by Rep. James Walkinshaw of Virginia, asked the Government Accountability Office to review how personnel and program reductions at CISA since early 2025 have affected federal network and critical infrastructure defense. The letter says the agency has lost roughly…
-
U.S. Bank says ransomware leak claim traces to a fourth-party incident
LockBit added U.S. Bancorp to its victim list Thursday morning and threatened to leak data in two weeks. U.S. Bancorp says the incident traces to a contractor of a third party, outside its own environment, and that there is no evidence its systems, networks, or data repositories were compromised…
-
More than 9,300 leaked AWS keys still authenticate, hundreds with full account control
Truffle Security tracked publicly exposed AWS access keys over four years and found 9,300-plus still valid. Of those, 817 belong to companies, 526 are root keys, 242 are IAM users with AdministratorAccess, and 130 live root keys sit on organization management accounts that control every member…
-
Malware reaches car head units through the built-in firmware updater
Kaspersky documented what it calls the first case of malware on an Android automotive head unit with an infection chain specific to that device type. The campaign abuses TWCore, the legitimate update app on head units running firmware from Chinese vendor DoFun, to install a dropper called…
-
Rust crate poisoning attributed to North Korean actors
Researchers have linked the compromise of the arrayref, internment, and append-only-vec crates to North Korean actors. The attacker used a compromised maintainer account to publish releases that added a typosquatted dependency whose build script downloaded and ran a remote payload during…
-
Fourteen trojanized npm packages drop an AI-assisted Linux backdoor
TrendAI found npm packages posing as calendar and streak utilities that bundle a binary, mark it executable, and launch it as a detached background process. The implant, RedC2 4.0, uses AI-assisted command and control. No install hook is needed: a single import anywhere in the dependency graph…
-
Cisco patches nine Crosswork and Secure Workload flaws, five at CVSS 10.0
Cisco released fixes for four vulnerabilities in Crosswork Data Gateway, Crosswork Network Controller, and Crosswork Planning that apply regardless of device configuration, including CVE-2026-20030 (SQL injection, CVSS 10.0), CVE-2026-20357 (missing authentication for a critical function, CVSS…
-
Phishing toolkit plants attacker-controlled passkeys that survive password resets
A toolkit called iAuthFlow V2, sold on a Russian-language forum for about $10,000 with add-on modules, runs an attacker-controlled browser that relays the victim's credentials and prompts to and from Google in real time. Once authentication completes, the toolkit shows a fake "Verification…
-
Microsoft Defender's own signed driver can be turned against security software at boot
Check Point Research disclosed a technique that abuses BTR.sys, Defender's Boot Time Removal Tool driver, to perform arbitrary kernel-level file and registry operations on Windows 7 through Windows 11 25H2. No software flaw is exploited and no outside driver is loaded, which means the technique…
-
Weekly infosec product releases: F5, Intezer, NETSCOUT, Tufin
NETSCOUT extended its Adaptive DDoS Protection to let service providers automatically detect and mitigate outbound DDoS traffic, meaning attack traffic originating from their own subscribers. F5 pushed enhancements to the F5 AI Gateway and folded it into the F5 AI Security Platform. Intezer and…
-
Defense contractors' CMMC confidence rises while ability to prove compliance falls
Average SPRS self-assessment scores climbed to a five-year high of +51, up from +33 in 2025. At the same time, the share of contractors extremely or very confident their score is accurate fell to 65 percent from 89 percent last year and 94 percent in 2024. In the companion survey, 96 percent said…
-
SickKids hospital breach exposes employee and job applicant data via third-party software flaw
Toronto's Hospital for Sick Children disclosed that a vulnerability in a third-party software application exposed personal information of current and former employees of SickKids, its Boomerang pediatric clinic, and the SickKids Foundation, plus job applicants. Clinical systems and patient records…
-
Attackers hide commands in FTP server banners to deliver two new RATs
SOCRadar documented a phishing-initiated chain that uses ZIP archives and LNK shortcut files to pull PowerShell from FTP server greeting banners, which act as dead-drop resolvers. The chain delivers two previously undocumented RATs. E4del is a Node.js RAT inside a digitally signed Electron app…
-
Microsoft patches maximum-severity Entra ID RCE already exploited in attacks
CVE-2026-69836 is a deserialization of untrusted data flaw in Microsoft Entra ID carrying a CVSS 10.0. It allowed an unauthenticated attacker to execute code over the network against Microsoft's cloud identity service, and Microsoft confirmed exploitation before disclosure. Microsoft says the flaw…
Also covered by The Hacker News
-
Critical isolated-vm flaw allows sandbox escape to host RCE
A type confusion bug in isolated-vm, a Node.js library for running untrusted JavaScript in a V8 isolate, allows V8 sandbox escape and control-flow hijacking of the host process. Tracked as GHSA-864f-rcv7-6rh4 with no CVE assigned, it affects all versions up to and including 7.0.0 and is fixed in…
-
Joint federal advisory: AI-generated exploitation scripts aimed at Siemens S7 PLCs
Five federal agencies warned that an unidentified threat actor is scanning for internet-exposed Siemens PLCs and building exploitation capability against them. The actor is using AI to generate Python exploitation scripts built on the snap7.dll and python-snap7 libraries, disguised as legitimate…
-
Operation CameraSwarm: single actor compromises more than 14,000 Dahua IP cameras
Hunt.io got into the operator's own infrastructure through an open HTTP directory the attacker left exposed, recovering 2,616 files across 234 subdirectories. The 35-day campaign ran June 17 to July 22 against Dahua cameras concentrated in Ukrainian, Russian, and CIS telecom netblocks. Methods…
-
Ransomware affiliate poses as a recovery firm to re-extort its own victims
An entity calling itself "Ransom Busters LTD" emails CEOs and IT leaders at freshly compromised organizations before the attack is public, claims to have infiltrated the criminal ransomware servers, and offers decryption keys and deletion of stolen data for $20,000 to $60,000. GRIT assesses with…
-
CISA orders federal patching of exploited TrueConf Server flaws used by Head Mare
CVE-2026-72529 is a missing authentication flaw letting an unauthenticated remote attacker execute arbitrary script via TCP port 4307. CVE-2026-72530 is a sandbox escape that breaks out to the underlying host through the same port. The hacktivist group Head Mare chained both to plant a web shell…
-
Microsoft ships 22 out-of-band cloud patches, four at maximum severity
Microsoft published 22 security updates covering Azure SQL Database, Azure Arc, Exchange Online, Azure Managed Instance for Apache Cassandra, and Entra ID. Four carry CVSS 10.0: CVE-2026-65816 and CVE-2026-69555 (unauthenticated remote privilege escalation in Azure Arc), CVE-2026-65801 (Exchange…
-
ToxicPanda 2.0 and GoldDigger expand Android banking fraud to 349 institutions across 16 countries
Zimperium zLabs documented an updated ToxicPanda (also called TgToxic) carrying 167 remote commands and a PIN-harvesting workflow targeting more than 140 banking and cryptocurrency applications. By abusing Android accessibility services the malware reads every UI element on screen and layers…
-
Forty malicious Firefox extensions impersonate Web3 wallets to steal seed phrases
Socket's threat research team confirmed 40 malicious Mozilla Firefox extensions that steal cryptocurrency wallet credentials while masquerading as OKX, Rabby Wallet, TronLink, and other Web3 products. They are part of a broader set of 77 add-ons sharing source code and infrastructure, tracked as…
-
NASA AIT-GUI flaw chain lets unauthenticated attackers issue spacecraft commands, CVSS 9.4
Cycode disclosed a chain of flaws in AIT-GUI, the browser-based operator console for NASA/JPL's open-source AMMOS Instrument Toolkit, rated 9.4 on CVSS v3.1 and affecting versions 2.5.1 and earlier, fixed in 2.5.2. The endpoints in question relay operator commands to a spacecraft and instrument…