Blahalla

WhatsApp Alerts 200 Users After Fake iOS App Installed Spyware; Italian Firm Faces Action

thehackernews.com · 6 months ago

WhatsApp alerted about 200 users after a fake iOS app infected with spyware was installed, with the vast majority of targets located in Italy. Threat actors used social engineering to trick users into installing a counterfeit WhatsApp app; affected accounts were logged out and users were advised to uninstall the malicious app and download the official WhatsApp app. WhatsApp said it is taking action against Asigint, an Italian subsidiary of spyware firm SIO accused of creating the counterfeit, and SIO has previously marketed surveillance tools to law enforcement and been linked to Android apps deploying a spyware family known as Spyrtacus.
NoVoice on Google Play Exploits 22 Flaws to Hit Millions of Android Users

gbhackers.com · 6 months ago

NoVoice is an Android rootkit campaign that hid in more than 50 Google Play apps and accumulated at least 2.3 million downloads before removal. It exploits 22 vulnerabilities patched between 2016 and 2021 to gain root on older and unpatched devices, uses polyglot images to hide encrypted APK payloads, and installs a system partition rootkit that replaces critical libraries to inject code into every app. Operators used 22 exploit binaries including IPv6 use after free and Mali GPU driver flaws to disable SELinux, enable silent app installation and data theft, and exfiltrate WhatsApp encrypted databases and identity keys to clone sessions; McAfee reported the campaign and Google removed the apps.
CISA Issues Alert on Chrome Zero-Day Under Active Exploitation

gbhackers.com · 6 months ago

There is a critical zero-day vulnerability affecting Google Chrome and other Chromium-based web browsers, tracked as CVE-2026-5281. The U.S. Cybersecurity and Infrastructure Security Agency issued an urgent warning and added the flaw to its Known Exploited Vulnerabilities catalog because attackers are actively exploiting it in real-world attacks.
Microsoft links Classic Outlook issue to email delivery problems

www.bleepingcomputer.com · 6 months ago

Microsoft is investigating a known issue that prevents some Classic Outlook users from sending emails via Outlook.com, with affected users receiving a non delivery report that reads you do not have the permission to send the message on behalf of the specified user and shows error code 0x80070005-0x0004dc-0x000524. The issue occurs more often when the Outlook.com account is an Outlook profile linked to another Exchange account and may also be triggered if the sender account has an Exchange Online mail contact with the same SMTP address. Microsoft recommends temporary workarounds including removing the Microsoft 365 account address book so the Outlook client does not check it when sending mail, hiding the Outlook.com contact from the tenant global address list, creating a new classic Outlook profile that only includes the affected account, or using the New Outlook client or Outlook on the web while a fix is sought.
Europe Pushes for a Gentler Internet for Children

www.nytimes.com · 6 months ago

false
Apple Releases iOS 18.7.7 Update to Defend Against DarkSword Exploit

gbhackers.com · 6 months ago

Apple expanded the rollout of iOS 18.7.7 and iPadOS 18.7.7 to defend users against a critical web based threat called the DarkSword exploit, originally released on March 24, 2026, and aggressively pushed to more devices via Automatic Updates on April 1, 2026. The update delivers mandatory web attack protections and patches over 15 significant security flaws across devices back to the iPhone XR and multiple iPad generations. Notable fixes include CVE-2026-20687, CVE-2026-20643, CVE-2026-28864, CVE-2026-28866, CVE-2026-28865, CVE-2025-43376, CVE-2026-28880 and CVE-2026-28878 addressing kernel use after free, WebKit cross origin and DNS leak issues, keychain and clipboard access problems, network authentication flaws, and permission errors that could let rogue apps enumerate installed software. Users are strongly advised to install the update immediately via Settings General Software Update or to allow the automatic installation.
Worried About A.I. Taking Your Job? That’s Not Very ‘Agentic’ of You.

www.nytimes.com · 6 months ago

false
Axios npm Supply Chain Breach: Microsoft Shares Mitigation Steps

gbhackers.com · 6 months ago

On March 31, 2026 two malicious Axios npm releases, versions 1.14.1 and 0.30.4, included a hidden dependency that executed at install time and has been attributed to the North Korean actor Sapphire Sleet. The installer contacted attacker command and control infrastructure on port 8000 at IP 142.11.206.73 and fetched platform specific remote access payloads, dropping a native macOS binary, a Windows PowerShell RAT that persists via a hidden Run registry key, and a Python loader for Linux. Recommended mitigations include rotating exposed secrets and credentials, downgrading to safe Axios versions 1.14.0 or 0.30.3 and removing malicious artifacts, pinning exact dependency versions and disabling automatic upgrades, reviewing CI CD install logs for the compromised versions or the rogue dependency plain crypto js, monitoring outbound connections to the identified C2 infrastructure, and adopting trusted publishing with OIDC to remove long lived npm tokens.
A.I. Could Change the World. But First It Is Changing Silicon Valley.

www.nytimes.com · 6 months ago

false
Cybersecurity in the age of instant software

www.csoonline.com · 6 months ago

AI is rapidly changing how software is written, deployed, and used, creating a future of ephemeral instant software that can be generated on demand and may coexist with traditional long lived applications. These AI capabilities are accelerating vulnerability discovery and exploitation, especially against open source libraries and low quality IoT and industrial software, while defenders can also use AI to find and patch flaws. Significant uncertainties remain about how well AIs will write secure code, how reliably they can create and deploy patches for legacy systems, and whether organizations will be able to install updates fast enough to close attack windows. If defenders can coordinate and automate patching and detection, networks could trend toward self healing, but attackers will shift to social engineering, credential theft, and manipulating defensive AIs through poisoning and prompt attacks.
Over 14,000 F5 BIG-IP APM instances still exposed to RCE attacks

www.bleepingcomputer.com · 6 months ago

Shadowserver reports that more than 14,000 F5 BIG-IP Access Policy Manager instances remain exposed online while attackers exploit a critical remote code execution vulnerability. The flaw was reclassified in March 2026 from a denial of service to remote code execution and is tracked as CVE-2025-53521. F5 warns that unauthenticated attackers can gain remote code execution on unpatched systems with access policies on a virtual server and has published indicators of compromise and guidance to check logs, disks, and terminal history and to rebuild affected devices from known good sources. The U S Cybersecurity and Infrastructure Security Agency added the vulnerability to its actively exploited list and ordered federal agencies to secure BIG-IP APM systems by the agency deadline.
FBI Warns Chinese Mobile Apps Could Expose User Data to Cyberattacks

gbhackers.com · 6 months ago

There is a public warning that foreign-developed mobile applications, particularly those from companies based in China, can expose user data because such apps may operate under national security laws that permit government access to data. Many apps request broad permissions that, once granted, can continuously collect sensitive personal information including contact lists and may store that data on servers in China for indefinite or unspecified periods. Some apps may embed malicious code capable of exploiting mobile operating system vulnerabilities to install backdoors, escalate privileges, and download additional malicious components, and apps from unofficial third-party stores carry significantly higher risk. To reduce exposure, users are advised to disable unnecessary permissions, download apps only from official stores, use strong authentication, keep devices updated with the latest patches, review privacy policies, and report suspected compromises to the Internet Crime Complaint Center.
New ZAP PTK Add-On Converts Browser Security Findings Into Native ZAP Alerts

gbhackers.com · 6 months ago

A new ZAP PTK Add-on 0.3.0, used with OWASP PenTest Kit 9.8.0, converts browser based security findings into native ZAP alerts and adds 142 PTK tagged alert types to the ZAP dashboard. There are three major features: native ZAP alerts for client side findings, custom rule selection between SAST, IAST, and DAST engines, and optional auto start scanning when the ZAP browser launches. The add on runs inside browsers launched by ZAP to monitor runtime signals and client side code, enabling detection of DOM based Cross Site Scripting, tainted inputs reaching sensitive DOM operations, unsafe innerHTML usage, and dangerous third party scripts that standard proxy scans miss.
Apple Expands iOS 18.7.7 Update to More Devices to Block DarkSword Exploit

thehackernews.com · 6 months ago

Apple expanded the availability of iOS 18.7.7 and iPadOS 18.7.7 on April 1, 2026 so devices with Automatic Updates can automatically receive security protections against the DarkSword exploit kit. The update now covers a broad range of iPhone models from iPhone XR through iPhone 16e and multiple iPad models, extending a patch first released on March 24, 2026 that had been limited to iPhone XS, iPhone XS Max, iPhone XR, and iPad 7th generation. DarkSword, disclosed by security researchers and active since July 2025, targets devices running iOS 18.4 through 18.7 via compromised websites to deploy backdoors and a data miner, and its leaked code on GitHub and use in attacks against targets in Saudi Arabia, Turkey, Malaysia, and Ukraine have raised concerns about wider exploitation.
one unrotated npm token. 100 million weekly downloads. 2 hours 54 minutes. this is what a real supply chain attack looks like.

www.codeant.ai · 6 months ago

On March 31, 2026 at 00:21 UTC a backdoored version of axios was published to the npm registry and for approximately two hours and fifty four minutes installing axios pulled a remote access trojan into developer machines, CI pipelines, and build servers. The attacker used a long lived classic npm access token to manually publish compromised releases that injected a newly published plain-crypto-js version 4.2.1 as a runtime dependency whose postinstall setup.js dropper executed automatically and used multi layer obfuscation, and the operation was pre staged with a clean plain-crypto-js version 4.2.0 and two attacker controlled npm accounts. Google Threat Intelligence attributed the campaign to UNC1069, npm removed and unpublished the malicious releases and placed a security placeholder on the malicious package, and the report notes that no CVE existed during the exposure window while recommending revoking classic tokens, requiring OIDC publishing, enabling two factor authentication, and publishing with provenance to reduce this class of supply chain risk.
AI Interview startup, Mercor Al breached via LiteLLM supply chain attack. Lapsus$ claims 4TB data breached including 211 GB candidate records and 3TB of video interviews

thecybersecguru.com · 6 months ago

On March 24, 2026, Mercor AI was reportedly breached in an incident linked to the hacking group Lapsus through a supply chain compromise of a LiteLLM package that was allegedly pulled by one of Mercor's AI agents. Attackers are said to have gained access to internal systems including Tailscale VPN credentials and to have exfiltrated roughly four terabytes of data, comprising 211 gigabytes of candidate records, 939 gigabytes of source code, and about three terabytes of video interviews and identity documents. Mercor acknowledged being among multiple companies affected and stated that its security team quickly contained the breach and began remediation efforts.
Cisco Smart Software Manager Flaw Allowed Arbitrary Command Execution

gbhackers.com · 6 months ago

A critical vulnerability tracked as CVE-2026-20160 in Cisco Smart Software Manager On-Prem allows an unauthenticated remote attacker to execute arbitrary commands with root privileges and carries a CVSS score of 9.8 out of 10. Releases from 9-202502 through 9-202510 are vulnerable while version 9-202601 contains the official fix, earlier releases before 9-202502 are not affected, and there are no available workarounds so administrators must apply the update. Cisco discovered the flaw internally during a support case and its Product Security Incident Response Team reports no evidence of public exploitation to date.
Critical PX4 Autopilot Vulnerability Let Attackers Gain Control of Drones

gbhackers.com · 6 months ago

CISA has issued a high priority alert about a critical PX4 Autopilot vulnerability tracked as CVE-2026-1579 with a CVSS v3 base score of 9.8. The flaw is a missing authentication for critical functions in PX4 version v1.16.0SITLlateststable that allows an attacker with access to the MAVLink interface to bypass security checks and execute arbitrary shell commands, enabling full takeover of the flight controller. Deployed across transportation systems, emergency services, and the defense industrial base, this vulnerability could enable stolen surveillance data, disruption of emergency response, or compromise of defense operations if exploited. Until a patch is available, operators are urged to restrict MAVLink interface access to trusted networks and users and to monitor the CISA advisory ICSA-26-090-02 for updates.
TA416 Broadens Europe Spy Campaign With Web Bugs and Malware

gbhackers.com · 6 months ago

TA416, a China aligned threat actor, resumed large scale espionage from mid 2025 against European government and diplomatic mailboxes and expanded operations to Middle Eastern diplomatic targets in March 2026. There are web bug reconnaissance steps embedded in phishing emails sent from freemail accounts and spoofed or compromised inboxes to profile victims, followed by evolving delivery chains that include fake Cloudflare Turnstile challenge pages impersonating Microsoft login flows, Microsoft Entra ID redirects, renamed MSBuild archives, ZIP smuggling with LNK files or CSPROJ downloaders, and DLL sideloading triads that load a customized PlugX backdoor. PlugX has been upgraded with stronger obfuscation and stealthy command and control techniques, including signed third party executables for DLL sideloading, API hashing, RC4 encrypted configuration blobs protected with rolling XOR, HTTP C2 using randomized URI paths and cookie values, and campaigns hosting archives on Google Drive or compromised SharePoint while using re registered domains fronted by Cloudflare to blend malicious infrastructure with legitimate traffic.
WhatsApp Attack Chain Delivers VBS, Cloud Payloads, MSI Backdoor

gbhackers.com · 6 months ago

A new malware campaign abuses WhatsApp messages to deliver malicious Visual Basic Script files to Windows users, enabling persistent remote access through unsigned MSI installers. The initial VBS attachments create hidden folders under ProgramData and copy legitimate Windows tools such as curl.exe and bitsadmin.exe into those locations while renaming them to misleading filenames, though PE metadata still identifies the original binaries. Secondary scripts fetched from cloud hosts including AWS S3, Tencent Cloud and Backblaze B2 attempt user account control bypasses and modify registry keys under HKLM Software Microsoft Win including ConsentPromptBehaviorAdmin to suppress prompts and escalate privileges. The campaign culminates in unsigned MSI packages with names like Setup.msi WinRAR.msi LinkPoint.msi and AnyDesk.msi that provide remote interactive access, and defenders are advised to block script hosts in untrusted locations enable cloud delivered protection run endpoint detection and response in block mode apply attack surface reduction rules and monitor cloud download patterns for suspicious activity.
Cisco source code stolen by ShinyHunters via Trivy supply-chain attack. AWS keys breached, 300+ repos cloned and more

thecybersecguru.com · 6 months ago

Attackers tied to ShinyHunters and TeamPCP leveraged credentials stolen in a March 19, 2026 compromise of the Trivy GitHub pipeline to breach Cisco's internal development environment and steal source code and cloud credentials. More than 300 private GitHub repositories were cloned, three million Salesforce records were exposed, and attackers leaked screenshots of AWS accounts and storage volumes, with stolen repositories reportedly including code for Cisco's AI Assistant, AI Defense algorithms, and blueprints for unreleased hardware and some customer repositories. Cisco rotated developer credentials, wiped and reimaged affected workstations, and isolated impacted AWS accounts while the threat actors issued an April 3, 2026 deadline threatening public release of the data.
Cisco Warns of Critical IMC Vulnerability Enabling Authentication Bypass

gbhackers.com · 6 months ago

There is a critical authentication bypass vulnerability tracked as CVE-2026-20093 with a CVSS base score of 9.8 that allows unauthenticated remote attackers to overwrite administrative passwords and gain full control of affected Cisco servers and preconfigured network appliances. The flaw is caused by the Integrated Management Controller failing to properly handle password change requests and can be exploited by sending a maliciously crafted HTTP request to the management interface. Affected hardware includes 5000 Series Enterprise Network Compute Systems, Catalyst 8300 Series Edge uCPE platforms, standalone UCS C-Series M5 and M6 rack servers, UCS E-Series M3 and M6 servers, and various preconfigured appliances such as APIC servers, Catalyst Center Appliances, Firewall Management Center Appliances, and Secure Network Analytics Appliances. There are no known workarounds or temporary mitigations, and administrators must urgently apply the vendor’s patched firmware or software updates to remediate the issue.
Remcos RAT Attack Uses Obfuscated Scripts, Trusted Windows Tools

gbhackers.com · 6 months ago

There is a Remcos RAT campaign that begins with a phishing email delivering a ZIP archive named MV MERKET COOPER SPECIFICATION.zip which contains a heavily obfuscated JavaScript file executed by Windows Script Host. The script creates ActiveX objects for command execution and HTTP requests, then downloads a PowerShell loader that uses Base64 and rotational XOR decryption to reconstruct and reflectively load a.NET assembly called ALTERNATE.dll entirely in memory. Operators also rebuild a portable executable in memory and abuse the trusted Windows utility aspnetcompiler.exe to host or proxy the malware, establishing persistent command and control connections, capturing keystrokes, and staging data for exfiltration. Defenders are advised to focus on detecting obfuscated script activity, PowerShell telemetry, living off the land binary abuse, and anomalous outbound connections rather than relying solely on file hashes and on access scanning.
Symantec DLP Agent Flaw Exposed Systems to Privilege Escalation Attacks

gbhackers.com · 6 months ago

A high-severity Local Privilege Escalation vulnerability in the Symantec Data Loss Prevention Agent for Windows allows low-privileged users to gain full SYSTEM-level control; it is tracked as CVE-2026-3991 and carries a CVSS score of 7.8. The flaw is caused by a hardcoded OpenSSL configuration file path that enables an attacker who can create the missing directory to drop a malicious configuration file and a DLL that the privileged edpa.exe process will load and execute with SYSTEM rights. Broadcom released an advisory and patches on March 30, 2026, and organizations should upgrade affected DLP installations to patched builds such as DLP 25.1 MP1, DLP 16.1 MP2, DLP 16.0 RU2 HF9, DLP 16.0 RU1 MP1 HF12, or DLP 16.0 MP2 HF15 to mitigate the issue.
Tools, um MCP-Server abzusichern

www.csoonline.com · 6 months ago

srcset="https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?quality=50&strip=all 7200w, https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?resize=300%2C168&quality=50&strip=all 300w, https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?resize=768%2C432&quality=50&strip=all 768w, https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?resize=1024%2C576&quality=50&strip=all 1024w, https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?resize=1536%2C864&quality=50&strip=all 1536w, https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?resize=2048%2C1152&quality=50&strip=all 2048w, https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?resize=1240%2C697&quality=50&strip=all 1240w, https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?resize=150%2C84&quality=50&strip=all 150w, https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?resize=854%2C480&quality=50&strip=all 854w, https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?resize=640%2C360&quality=50&strip=all 640w, https://b2b-contenthub.com/wp-content/uploads/2025/11/Gorodenkoffshutterstock232495234716z9.jpg?resize=444%2C250&quality=50&strip=all 444w" width="1024" height="576" sizes="auto, (max-width: 1024px) 100vw, 1024px" Unabhängig davon, welche MCP-Server Unternehmen wofür einsetzen, “Unsicherheiten” sollten dabei außenvorbleiben. Gorodenkoff | shutterstock.com Model Context Protocol ( MCP ) verbindet KI-Agenten mit Datenquellen und erfreut sich im Unternehmensumfeld wachsender Beliebtheit. Allerdings ist auch MCP nicht frei von Sicherheitslücken, wie entsprechende Entdeckungen, etwa beim SaaS-Anbieter Asana oder dem IT-Riesen Atlassian gezeigt haben. Inzwischen hat sich jedoch einiges in Sachen MCP-Sicherheit getan. Einerseits wurden mit Blick auf das Kernprotokoll etliche Fortschritte erzielt. Beispielsweise in Form von Support für OAuth sowie für Authentifizierungs-Server von Drittanbietern und Identity-Management-Systeme. Darüber hinaus wurde inzwischen auch eine offizielle MCP Registry geschaffen, die einen Überblick über sichere, öffentlich verfügbare MCP-Server bietet. Dennoch bestehen weiterhin Sicherheitslücken, die sich für diverse Cyberschandtaten ausnutzen lassen, Prompt Injection, Tool Poisoning, Token-Diebstahl, Server-übergreifende Attacken oder manipulierte Messages sind nur einige von vielen Beispielen. Mit anderen Worten: Unternehmen, die sich beim Aufbau von Agentic-AI-Systemen einen Wettbewerbsvorteil verschaffen wollen, müssen erhebliche Anstrengungen unternehmen, um zu gewährleisten, dass sensible Daten nicht nach außen dringen. Glücklicherweise gibt es diverse Tools, die dabei Unterstützung versprechen. In diesem Artikel lesen Sie: was Security-Tools für MCP leisten sollten, und welche Angebote in diesem Bereich interessant sind. Das sollten MCP-Sicherheitslösungen können Die Gefahr von Datenlecks, Prompt Injections und weiteren Sicherheitsbedrohungen besteht unabhängig davon, ob Unternehmen: ihre eigenen KI-Agenten mit MCP-Servern von Drittanbietern, ihre eigenen MCP-Server mit Drittanbieter-Agenten, oder ihre eigenen Server mit den eigenen Agenten verbinden. Soll heißen: Unternehmen müssen in jedem Fall Autorisierungen und Berechtigungen überprüfen, detaillierte Zugriffskontrollen implementieren und alles protokollieren. Daraus ergeben sich auch die Anforderungen für MCP-Sicherheitslösungen. Diese sollten bieten: MCP-Servererkennung. Für Mitarbeiter eines Unternehmens ist es einfach, MCP-Server herunterzuladen und zu nutzen. Mit Scan-Services für MCP-Server können Unternehmen sämtliche Instanzen von Schatten-MCP-Servern in ihrer Umgebung finden. Laufzeitschutz. KI-Agenten kommunizieren mit MCP-Servern in natürlicher Sprache. MCP-Sicherheits-Tools sollten deshalb in der Lage sein, diese Kommunikation auf Sicherheitsprobleme wie Prompt Injections hin zu überwachen. Authentifizierungs- und Zugriffskontrollen. Das MCP-Protokoll unterstützt inzwischen OAuth, aber das ist nur ein erster Schritt. Für zusätzliche Sicherheit empfehlen sich Tools mit integrierten Kontroll-Frameworks für Zero Trust und Least Privilege. Logging und Observability. Tools und Plattformen sollten zudem die Möglichkeit bieten, MCP-Protokolle zu sammeln, Sicherheitsteams über Richtlinienverstöße zu informieren, Compliance-Daten zu erfassen oder Protokolle in die bestehende Sicherheitsinfrastruktur einzuspeisen. MCP-Security-Angebote Im Folgenden haben wir die Anbieter von MCP-Security-Tools in drei Kategorien aufgeteilt. Diese Aufstellung erhebt keinen Anspruch auf Vollständigkeit. Hyperscaler Für Unternehmen, die sich vollständig auf eine bestimmte Cloud-Plattform verlassen, bieten die MCP-Tools des jeweiligen Hyperscalers einen einfachen Einstieg. Amazon Web Services (AWS) hat Mitte 2025 seine eigene agentenbasierte KI-Plattform eingeführt. Amazon Bedrock AgentCore umfasst ein Gateway, das mehrere Protokolle unterstützt (darunter auch MCP), ein Identity-Management-System sowie Observability. Microsoft bietet einen grundlegenden Azure-MCP-Server an, inklusive Support für Azure Key Vault. Darüber hinaus unterstützen auch Azure AI Foundry Agent Service und Azure API Management das Model Context Protocol. Zudem bietet Microsoft mit dem Agent Framework auch ein Open-Source-Entwicklungskit, das sowohl MCP als auch Agent2Agent unterstützt und beispielsweise Schutz vor Prompt Injections verspricht. Google Cloud kündigte Anfang 2025 seine MCP Toolbox für Datenbanken an, inklusive integrierter Authentifizierung und Observability. Außerdem hat der Hyperscaler auch eine Referenzarchitektur veröffentlicht, um MCP-Server auf seiner Cloud-Plattform abzusichern. Große Plattformanbieter Der IT-Dienstleister Cloudflare hat mit MCP Server Portals ein Tool veröffentlicht, mit dem Unternehmen MCP-Verbindungen zentralisiert absichern und überwachen können. Die Funktion ist Bestandteil der Cloudflare-One-Plattform. Palo Alto Networks hat mit Blick auf MCP-Sicherheit mehrere Eisen im Feuer. Mit Prisma AIRS hat das Unternehmen einen eigenen, intermediären MCP-Server veröffentlicht. Dieser sitzt zwischen den KI-Agenten und dem eigentlichen MCP-Server und erkennt schadhafte Inhalte und Daten. Das Tool MCP Security ist hingegen Bestandteil von Cortex Cloud WAAS und überprüft die MCP-Kommunikation an der Netzwerkgrenze auf bösartige Aktivitäten. SentinelOne gewährt mit seiner Singularity Platform ebenfalls Einblick in die MCP-Interaktionskette und bietet zum Beispiel Warnmeldungen und automatisierte Incident Response für MCP-Server auf lokaler oder Remote-Ebene. Daneben hat auch Broadcom MCP-Sicherheitsfunktionen für VMware Cloud Foundation angekündigt, die künftig mehr Sicherheit für agentenbasierte Workflows gewährleisten sollen. Startups Die Plattform von Acuvity verspricht, MCP-Server umfassend abzusichern. Dafür sorgt laut dem Anbieter eine Kombination aus Least-Privilege-Execution, unveränderlichen Laufzeiten, kontinuierlichen Schwachstellenscans, Authentifizierung und Bedrohungserkennung. Das API-Security-Startup Akto hat eine MCP-Security-Plattform im Angebot. Sie umfasst ein Discovery Tool, um MCP-Server in Unternehmensumgebungen zu identifizieren, Security-Testing-Werkzeuge sowie Monitoring- und Threat-Detection-Funktionen. Invariant Labs bietet mit MCP-Scan ein quelloffenes Tool, das die statische Analyse und Echtzeitüberwachung von MCP-Servern ermöglicht. Mit Guardrails hat das Startup auch ein kommerzielles Produkt im Angebot. Dabei handelt es sich um einen Proxy. Der zwischen KI-Agenten und MCP-Servern sitzt und vor Security-Risiken schützen soll. Das Tool befähigt Anwender außerdem dazu, Richtlinien aufzusetzen. Die AI Security Fabric-Plattform von Javelin addressiert ebenfalls das Thema MCP-Sicherheit. Etwa mit Funktionen wie MCP-Server auf Risiken zu scannen oder Datenanfragen zu überprüfen. Lasso Security stellt ein Open-Source- MCP-Gateway zur Verfügung, das die Konfiguration und das Lebenszyklusmanagement von MCP-Servern ermöglicht und Messages um sensible Informationen bereinigt. (fm) Sie wollen weitere interessante Beiträge rund um das Thema IT-Sicherheit lesen? Unser kostenloser Newsletter liefert Ihnen alles, was Sicherheitsentscheider und -experten wissen sollten, direkt in Ihre Inbox.
Smashing Security podcast #461: This man hid $400 million in a fishing rod. Then it vanished

grahamcluley.com · 6 months ago

An Irish cannabis grower and gyrocopter pilot who invested drug money in Bitcoin in 2011 reportedly hid access codes in a fishing rod case that went missing, yet one of his previously frozen wallets recently moved thirty five million dollars after someone presented identification to authorize the transfer. A data breach at Ajax Football Club that was said to affect a few hundred fans is now believed to have exposed personal details of three hundred thousand supporters and included flaws that could allow attackers to steal match tickets and remove people from stadium ban lists. Both stories were discussed on episode four hundred sixty one of the Smashing Security podcast with guest Danny Palmer.
New CrystalRAT malware adds RAT, stealer and prankware features

www.bleepingcomputer.com · 6 months ago

A new malware as a service called CrystalX, sometimes referred to as CrystalRAT, is being promoted on Telegram and YouTube with a tiered subscription model and an automated builder that allows customization and anti analysis options. The toolkit offers a user friendly control panel and builder, communicates with command and control over WebSocket, delivers zlib compressed payloads encrypted with ChaCha20, and includes an infostealer targeting Chromium based browsers and desktop apps, a remote access module with VNC, microphone and camera capture, a real time keylogger, and a clipboard clipper. It also contains extensive prankware commands that can change wallpapers, alter display orientation, remap input, hide system UI elements, force shutdowns, and show fake notifications which can annoy or distract victims while data theft runs in the background.
Cisco Evolved Programmable Network Manager Improper Authorization Vulnerability

sec.cloudapps.cisco.com · 6 months ago

A vulnerability in the web based management interface of Cisco Evolved Programmable Network Manager allows an authenticated remote attacker with low privileges to access sensitive information because of improper authorization checks on a REST API endpoint. An attacker can query the endpoint to view session information of active EPNM users including administrators, which could result in the affected device being compromised. Cisco has released software updates to address the issue, no workarounds are available, the security impact rating is High, CVE-2026-20155.
Cisco Smart Software Manager On-Prem Arbitrary Command Execution Vulnerability

sec.cloudapps.cisco.com · 6 months ago

A vulnerability in Cisco Smart Software Manager On-Prem allows an unauthenticated remote attacker to execute arbitrary commands on the underlying operating system of an affected host. The issue is caused by unintentional exposure of an internal service and can be exploited by sending a crafted request to that service API, potentially resulting in command execution with root privileges. Cisco has released software updates to address the vulnerability and reports that there are no workarounds. The vulnerability is tracked as CVE-2026-20160 and has a security impact rating of critical.
Cisco Integrated Management Controller Authentication Bypass Vulnerability

sec.cloudapps.cisco.com · 6 months ago

A vulnerability in the change password functionality of Cisco Integrated Management Controller could allow an unauthenticated remote attacker to bypass authentication and gain access as Admin. An attacker could exploit the flaw by sending a crafted HTTP request to an affected device, then alter the passwords of any user on the system, including an Admin user, to obtain access. Software updates have been released to address the issue, no workarounds exist, the security impact rating is Critical, CVE CVE-2026-20093.
Cisco Smart Software Manager On-Prem Privilege Escalation Vulnerability

sec.cloudapps.cisco.com · 6 months ago

A high severity vulnerability in the web interface of Cisco Smart Software Manager On-Prem allows an authenticated remote attacker with at least System User credentials to elevate privileges to administrative. The issue is caused by improper transmission of sensitive user information, enabling an attacker to send a crafted message to an affected host and retrieve session credentials from subsequent status messages. Only users who are currently logged in through the web interface are exposed and SSH sessions are not affected. Cisco has released software updates with no available workarounds and the flaw is tracked as CVE-2026-20151.
Goodbye ‘Geeky Hunk’? Gmail Users Can Now Change Their Usernames.

www.nytimes.com · 6 months ago

Google is rolling out a change in the United States that allows Gmail users to replace their @gmail.com address with a new @gmail.com address without losing emails, photos, Drive files, subscriptions, or other account data. The previous address becomes an alternate alias that still receives mail, and users can sign in to Google services with either the old or new address. Users can revert to a prior address but are limited to changing their Gmail address once every 12 months and up to three times in total, and old addresses cannot be reused by other accounts.
Ransomware Will Hit Hospitals. Rehearsals Are Key to Defense

www.darkreading.com · 6 months ago

There are frequent ransomware attacks against hospitals and rehearsing downtime procedures is essential because real incidents under pressure differ from training. Recommendations include practicing offline patient care with pen and paper, implementing redundant identity verification such as two-person high-risk confirmation and pre-validated paper medication administration records, and running tabletop exercises that include frontline staff, since downtime playbooks help mainly for short outages and organizations must rehearse partial and gray-zone failures as well. Hospitals should map where identity, information, and execution depend on digital systems, be cautious about risks from unapproved AI tools, and recognize that preparation determines whether a situation escalates or stabilizes.
Apple expands iOS 18 updates to more iPhones to block DarkSword attacks

www.bleepingcomputer.com · 6 months ago

There are new iOS 18.7.7 security updates made available on April 1, 2026 that expand protections to more iPhones still running iOS 18 against web attacks called DarkSword. There are six vulnerabilities exploited by DarkSword tracked as CVE-2025-31277, CVE-2025-43529, CVE-2026-20700, CVE-2025-14174, CVE-2025-43510, and CVE-2025-43520. There are reports that the exploit kit, which first targeted iOS 18.4 through 18.7, was used widely by commercial and state linked actors including PARS Defense, UNC6748, and a suspected Russian espionage group UNC6353, and that attackers deployed JavaScript infostealers and backdoors named GhostBlade, GhostKnife, and GhostSaber. There are many models now eligible for iOS 18.7.7 including iPhone XR through iPhone 16 and several iPad models, and devices with Automatic Updates enabled will automatically receive the protections.
Hackers exploit TrueConf zero-day to push malicious software updates

www.bleepingcomputer.com · 6 months ago

There is a zero day in TrueConf tracked as CVE-2026-3502 that stems from a missing integrity check in the updater and allows an on premises TrueConf server to replace legitimate updates with malicious executables, enabling code execution on all connected endpoints. The flaw affects TrueConf versions 8.1.0 through 8.5.2 and was patched in version 8.5.3 released in March 2026. A campaign named TrueChaos has exploited the vulnerability to push weaponized updates to centrally managed government TrueConf servers in Southeast Asia, deploying a Havoc command and control implant and using DLL sideloading, reconnaissance commands, and a UAC bypass via iscicpl.exe. Indicators of compromise include poweriso.exe, 7z dash x64 dot dll, iscsiexe dot dll and artifacts such as AppData Roaming Adobe update dot 7z, and analysts assess with moderate confidence the activity is linked to a Chinese nexus threat actor using Alibaba Cloud and Tencent hosting.
Mitigating the Axios npm supply chain compromise

www.microsoft.com · 6 months ago

On March 31, 2026 two malicious npm releases of Axios identified as versions 1.14.1 and 0.30.4 included an injected dependency that fetched a second stage remote access trojan during installation. Microsoft Threat Intelligence attributed the compromise and associated infrastructure to the North Korean state actor known as Sapphire Sleet and confirmed the packages used a post install lifecycle hook in plain-crypto-js version 4.2.1 to execute setup.js and deploy OS specific RAT payloads to Windows macOS and Linux. Users are advised to downgrade to safe versions 1.14.0 or 0.30.3 rotate exposed secrets disable auto updates and follow mitigation guidance from security vendors to contain and remediate the compromise.
Elon Musk’s SpaceX files to go public, seeking massive market debut

www.washingtonpost.com · 6 months ago

SpaceX has filed to go public in a confidential IPO filing that could lead to a June 2026 market debut and a valuation as high as 1.75 trillion, potentially making it the largest offering in history. The company, valued at more than 1 trillion, has grown since its 2002 founding by deploying thousands of satellites, pioneering reusable rockets, winning NASA contracts, and absorbing Elon Musk's xAI and the X social platform. SpaceX says it is prioritizing Moon missions, data centers in space and satellite internet even as it faces scrutiny about readiness for a crewed lunar mission following public disputes with government officials.
WhatsApp Whistle-Blower Suit Is Dismissed

www.nytimes.com · 6 months ago

NULL
New EvilTokens service fuels Microsoft device code phishing attacks

www.bleepingcomputer.com · 6 months ago

There is a new malicious kit called EvilTokens that integrates device code phishing to hijack Microsoft accounts, is sold via Telegram, and its author says they plan to add Gmail and Okta phishing support. The attacks abuse the OAuth 2.0 device authorization flow by tricking victims through documents containing QR codes or links into authorizing a malicious device, enabling attackers to capture short lived access tokens and refresh tokens for persistent access to email, files, Teams, and single sign on impersonation. Researchers observed campaigns targeting finance, HR, logistics, and sales personnel across multiple countries including the United States, Canada, France, Australia, India, Switzerland, and the United Arab Emirates, and published indicators of compromise and YARA rules to help defenders block EvilTokens.
Elon Musk’s SpaceX Files to Go Public, Setting Stage for Huge I.P.O.

www.nytimes.com · 6 months ago

false
Securing the open source supply chain across GitHub

github.blog · 6 months ago

Recent attacks on the open source supply chain focus on exfiltrating secrets such as API keys to publish malicious packages and propagate through projects, often by compromising GitHub Actions workflows. Immediate prevention steps include enabling CodeQL to inspect workflows, avoiding workflows triggered by pull request target events, pinning third-party Actions to full-length commit SHAs, watching for script injection from user-submitted content, and replacing secrets with OpenID Connect workload identity tokens. There are platform-level measures including trusted publishing integrated with package repositories such as npm, PyPI, NuGet, RubyGems, and Crates, continuous malware scanning of npm packages, publishing compromised dependency information in an advisory database, and using Dependabot to notify maintainers. In response to the late 2025 Shai-Hulud attacks GitHub accelerated rollout of trusted publishing and is revisiting the GitHub Actions security roadmap to add secure defaults, policy controls, and CI CD observability.
Welcome to the war party: The new Trump-tech alliance cashes in

www.washingtonpost.com · 6 months ago

There is a growing realignment between Silicon Valley investors and the Trump administration as venture-backed startups are attracting large Pentagon funding, including a $400 million Pentagon investment in MP Materials, a $900 million Navy commitment to Hadrian, and Palantir’s Maven being designated a program of record. The Hill and Valley Forum has evolved into a major Washington convening where former tech executives now in government, such as David Sacks, Emil Michael and Jacob Helberg, are pushing policies and fast-tracked acquisitions that shift defense spending toward nimble startups while industry and media allies celebrate the dealmaking. Critics and some attendees warn that many of these companies will fail, that private profit and public power are increasingly intertwined, and that the current moment may be temporary rather than a lasting transformation of the defense industrial base.
LatAm's Self-Taught Cyber Talent Overlooked Amid Cyberattack Glut

www.darkreading.com · 6 months ago

A jobs study based on a survey of 605 Latin American cybersecurity professionals finds that Latin America faces forty percent more cyberattacks than the global average and that the region's threat landscape is outpacing available defensive talent. Seventy percent of respondents reported acquiring skills through informal learning such as online courses and on the job experience, only forty four percent held a university degree, fifty three percent held at least one certification, thirty five percent had under three years of experience, and many maintain a second cybersecurity related occupation. Employers are urged to expand hiring beyond traditional credentials, retune experience and compensation requirements, and emphasize flexibility, wellbeing, training, and job stability to tap underutilized local talent and reduce hiring gaps.
Threat Brief: Widespread Impact of the Axios Supply Chain Attack

unit42.paloaltonetworks.com · 6 months ago

A supply chain attack targeted the Axios JavaScript library after a maintainer's npm account was hijacked, leading to malicious releases v1.14.1 and v0.30.4 that injected a dependency named plain-crypto-js which acted as a cross platform remote access Trojan affecting Windows, macOS and Linux. The compromise used npm's postinstall lifecycle hook to run an obfuscated Node.js dropper that fetched platform specific payloads and delivered macOS, Windows and Linux RATs implemented in C++, PowerShell and Python respectively. All variants share an identical command and control protocol, beacon regularly with base64 encoded JSON over HTTP POST, accept commands such as kill, runscript, peinject and rundir, and establish persistence while performing reconnaissance and anti forensic cleanup by deleting the dropper and restoring a clean decoy package file. Analysis shows overlap with WAVESHAPER and links to previously reported DPRK operations, and the campaign has affected organizations across sectors including business services, finance, higher education, media and medical equipment.
A.I. Companies Shatter Fund-Raising Records, as Boom Accelerates

www.nytimes.com · 6 months ago

false
'NoVoice' Android malware on Google Play infected 2.3 million devices

www.bleepingcomputer.com · 6 months ago

There is a new Android malware named NoVoice found on Google Play hidden in more than 50 apps that were downloaded at least 2.3 million times. The infected apps, including cleaners, image galleries, and games, requested no suspicious permissions and used steganography to hide an encrypted APK inside PNG images which is extracted and loaded in memory while intermediate files are wiped. The malware probes device hardware, kernel and Android patch level, then downloads device-specific exploit modules to gain root using as many as 22 exploits including use-after-free kernel bugs and Mali GPU driver flaws patched between 2016 and May 2021, installs a persistent rootkit that can survive factory resets, and exfiltrates WhatsApp encryption databases and keys to allow session cloning. The malicious apps have been removed from Google Play after McAfee reported them, but previously infected users should treat their devices and data as compromised and upgrade to devices with security patches newer than May 2021.
Vim and GNU Emacs: Claude Code helpfully found zero-day exploits for both

www.csoonline.com · 6 months ago

Researcher Hung Nguyen used Anthropic's Claude Code to quickly discover zero-day remote code execution vulnerabilities in the source code of the developer editors Vim and GNU Emacs. Claude Code found missing security checks in Vim's tabpanel sidebar and in the autocmdadd function and produced a proof of concept that bypassed Vim's sandbox to execute arbitrary commands when a user opens a crafted file; Vim maintainers fixed the issue as CVE-2026-34714 with a CVSS score of 9.2 in version 9.2.0272. Claude Code also identified an Emacs vulnerability dating back to 2018 where opening a file inside a directory containing a crafted dot git folder can trigger attacker controlled commands, affecting versions 30.2 and 31.0.50, and Emacs maintainers have not assigned a CVE or issued a patch while suggesting manual mitigations, illustrating how rapidly large language models can find and weaponize long hidden flaws.
One of Apple’s First Employees Looks Back at 50 Years

www.nytimes.com · 6 months ago

Chris Espinosa joined Apple in 1976 at age 14, riding a moped to demonstrate computers being built in Steve Jobs’s childhood home. He wrote BASIC code in the company’s earliest days and later worked in engineering and publications as Apple expanded into a global technology company. Espinosa has remained at Apple for fifty years and was a featured participant in the company’s fiftieth-anniversary events, widely recognized as the company’s longest-serving employee.
Is “Hackback” Official US Cybersecurity Strategy?

www.schneier.com · 6 months ago

The 2026 US Cyber Strategy for America, issued in March 2026, largely continues longstanding White House cyber policy but adopts a notably more aggressive tone. It includes a provision to "unleash the private sector by creating incentives to identify and disrupt adversary networks and scale our national capabilities," language that can be read as authorizing private-sector offensive cyber operations. Security commentators warn this approach moves toward hackback, risks vigilantism, undermines due process, and could dangerously escalate conflicts.
Cyberattacks Intensify Pressure on Latin American Governments

www.darkreading.com · 6 months ago

There are an estimated 3,050 cyberattacks per week against organizations in Latin America in March, compared with about 2,000 per week globally, and government agencies faced nearly 4,200 attacks per week. Attackers range from financially motivated criminals and ransomware affiliates to state-related actors and hacktivists, with phishing and information-stealers driving a surge in stolen credentials and email remaining the primary delivery vector. High-profile March incidents included AI-assisted compromises of at least nine Mexican government agencies likely exposing more than 195 million identities and tax records, more than 23 million attacks and probes against Colombia’s national health regulator, and a disruptive intrusion that halted Puerto Rico’s transportation department from issuing driver's licenses. Analysts warn that legacy systems, uneven patching, and a regional shortage of cybersecurity professionals increase risk and recommend prioritizing email protections, external attack-surface scanning, real-time visibility, and remediation of the most exploitable exposures.