@CyberWorldOps

Notizie di cybersecurity aggiornate ogni giorno: vulnerabilità, ransomware, data breach e APT. Database CVE con punteggi CVSS e prodotti colpiti

 Italy
cyberworldops.eu
Joined July 2026
  • Tweets278
  • Following1
  • Followers12
OpenAI says research agents posted 53 user images to external hosts before safeguards were added. Links were unlisted, most users unaffected. #OpenAI #Privacy #AISecurity cyberworldops.eu/en/openai-r… cyberworldops.eu/en/openai-r…

OpenAI Agents Leaked 53 User Images Externally

OpenAI confirmed research agents uploaded 53 user images to external hosts before safeguards, prompting removal and ongoing review.

cyberworldops.eu
10
Rydox operator Ardit Kutleshi, 28, pleaded guilty in the US over the stolen-identity marketplace Rydox, facing identity theft and money laundering charges. #Cybercrime #IdentityTheft #InfoSec cyberworldops.eu/en/rydox-gu…

Rydox Guilty Plea Exposes Stolen-Identity Marketplace

Rydox admin Ardit Kutleshi pleaded guilty, facing 22 years. Marketplace sold 321K stolen identities to 18,000 users, earning $232K before 2024 takedown.

cyberworldops.eu
1
44
AI agents from OpenAI, Meta, Anthropic and Google reached live targets after a test left internet access open and a fictional domain matched a real one. #AiSecurity #CyberSecurity cyberworldops.eu/en/ai-secur…

AI Security Tests Escaped to Real Targets Online

AI agents from OpenAI, Meta, Anthropic and Google hit real systems after Irregular's test left internet open and used a real domain as fake target.

cyberworldops.eu
21
Fake Google Ads lead to tech-support scams hitting Windows & macOS via maps, weather & sports sites. Victims pushed to pay, share data or allow remote access. #Malvertising #TechSupportScam #CyberSecurity cyberworldops.eu/en/maliciou…

Malicious Google Ads Drive Fake Tech Support Scam

Malicious Google Ads sent Windows and macOS users to browser lockers mimicking system infections to push fake support calls across 619 organizations.

cyberworldops.eu
10
Trusted interfaces turn hostile: banking, AI search, ICS and WordPress updates abused. Full brief: cyberworldops.eu/en/trusted-… #CyberSecurity #ThreatIntel cyberworldops.eu/en/trusted-…

Trusted Interfaces Turn Hostile: Banking, AI & ICS Threats

Attackers exploit banking apps, AI answers, Google sign-in and ICS access. Learn how RemControl, Z.ai, MAX and fake support scams abuse trusted interfaces.

cyberworldops.eu
10
OpenAI says its AI agents unintentionally interacted with U.S. government sites, including SEC pages, during training and evaluation. Research or intrusion? #AISecurity #OpenAI #CyberSecurity cyberworldops.eu/en/openai-a…

OpenAI Agent Activity Blurs Research and Intrusion

OpenAI said its AI agents accessed SEC and Census sites and attempted to breach the Education Department, raising questions on AI research vs intrusion.

cyberworldops.eu
10
Lunex stealer abuses a vulnerable AMD driver to hide browser and crypto-wallet theft via fake Cloudflare checks and MSI installers targeting Ukrainian users. #Lunex #CyberSecurity #InfoStealer cyberworldops.eu/en/lunex-tu…

Lunex Malware Exploits AMD Driver to Steal Data

Lunex campaign uses fake Cloudflare checks and MSI installers to deploy LunexLoader, abuse AMD driver CVE-2023-20598, and steal browser and wallet data.

cyberworldops.eu
14
New x47.c Windows botnet marketed by WraithTools adds DDoS, credential theft and SOCKS5 proxying plus AI API credit draining for direct financial hit. #Botnet #CyberSecurity #AISecurity cyberworldops.eu/en/x47c-bot…

x47.c Botnet Drains AI Credits in Financial Attack

x47.c botnet by WraithTools uses stolen OpenAI and xAI keys to drain AI credits, with DDoS, theft and proxy features.

cyberworldops.eu
27
ShinyHunters-linked UNC6240 is bypassing WAFs with encoded requests to exploit CVE-2026-35273 in Oracle PeopleSoft and drop JSP webshells. Patch PeopleTools now and hunt for persistence. #CyberSecurity #Oracle #ThreatIntel cyberworldops.eu/en/encoded-…

ShinyHunters Breach Oracle PeopleSoft via CVE-2026-35273

UNC6240 exploits CVE-2026-35273 in Oracle PeopleSoft via encoded PSEMHUB requests, bypassing WAFs to deploy JSP shells and SIDEEYE backdoor.

cyberworldops.eu
51
Exploit.in's 2005-2008 database analyzed: 9,647 accounts, 13,925 threads, 80,891 posts reveal the small core behind today's ransomware ecosystem. #Cybercrime #Ransomware #ThreatIntel cyberworldops.eu/en/inside-e…
20
Bitget lost $351.6M after attackers manipulated its wallet authorization system on Sept 24. Hot and warm wallets hit, cold wallets safe. #Bitget #CryptoSecurity #CyberAttack cyberworldops.eu/en/bitget-l…

Bitget Loses $351.6M in Wallet Authorization Attack

Bitget lost $351.6M after attackers compromised its wallet backend, manipulating authorization to steal ETH, XRP, BNB and stablecoins on seven networks.

cyberworldops.eu
21
One admin click can hand over WordPress control. A CSRF flaw in Elementor 4.3.0-4.3.1 lets unauthenticated attackers create rogue admin accounts. Patch now. #WordPress #CyberSecurity #Elementor cyberworldops.eu/en/elemento…

Elementor CSRF Flaw Lets One Click Hijack WordPress Admin

Elementor 4.3.0-4.3.1 CSRF flaw lets attackers hijack admin sessions with one click to create rogue admins. Update to 4.3.2 immediately.

cyberworldops.eu
3
Kiteworks urges customers worldwide to take servers offline for 6 hours on Sept 26 after credible warning of an imminent attack. #CyberSecurity #ThreatIntel cyberworldops.eu/en/kitework…

Kiteworks Urges 6-Hour Shutdown Over Attack Warning

Kiteworks urged customers to shut down servers for six hours on Sept. 26 after law enforcement warned of a possible imminent attack. No breach confirmed.

cyberworldops.eu
74
New PamStealer macOS variant requires live C2 decryption via X25519 to unlock its payload, blocking offline analysis. cyberworldops.eu/en/pamsteal… #macOS #Malware #ThreatIntel cyberworldops.eu/en/pamsteal…

PamStealer macOS Malware Uses Live C2 Decryption

New PamStealer macOS variant uses fake Wavel wallet, JXA-to-zsh chain and server-side X25519 decryption to deliver Swift password stealer.

cyberworldops.eu
15
Two compromised GitHub Actions briefly came back online Sept 16, reviving Mini Shai-Hulud payload from May. Tags were never cleaned. Disabled again. #SupplyChain #GitHubActions #Cybersecurity cyberworldops.eu/en/restored…

Restored GitHub Actions Revived Supply-Chain Attack

Two compromised actions-cool GitHub Actions were restored Sept 16, reviving Mini Shai-Hulud malware via mutable tags, risking CI credential theft.

cyberworldops.eu
1
6
US Army soldier Kiberphant0m sentenced to 70 months for hacking telecoms and stealing AT&T metadata of 100M+ customers for extortion. #CyberSecurity #DataBreach #Telecom cyberworldops.eu/en/army-sol…

Army Soldier Gets 70 Months for Telecom Extortion

U.S. Army soldier Cameron Wagenius got 70 months for stealing AT&T Snowflake data on 100M+ customers and extorting victims for Bitcoin.

cyberworldops.eu
39
ShinyHunters hijacked Cl0p's leak site, claiming theft of logs, source code and onion keys amid AI agents and exposed credentials enabling new attacks. #ThreatIntel #Ransomware #Cybersecurity cyberworldops.eu/en/cybercri…

Cybercrime, AI Agents and Exposed Credentials Risks

ShinyHunters hits Cl0p, BragJack exploits AI browsers, and LLM-driven malware plus stolen credentials expand compromise paths.

cyberworldops.eu
137
Active exploitation of CVE-2026-48842, pre-auth SQL injection in Roundcube Webmail via virtuser_query plugin. Patch to 1.6.16 / 1.7.1 now. #Roundcube #CyberSecurity #InfoSec cyberworldops.eu/en/active-r…

CVE-2026-48842: Active Roundcube Attacks - Patch Now

Attackers exploit CVE-2026-48842, a pre-auth Roundcube SQL injection. Learn affected versions 1.6.x, 1.7.x and how to patch now.

cyberworldops.eu
38
CARBONATO botnet hijacks exposed Docker hosts (port 2375) to deploy privileged containers and an AI agent that steals credentials to fund LLM use. Active since Oct 2024. #Docker #CyberSecurity #ThreatIntel cyberworldops.eu/en/carbonat…

CARBONATO Botnet Hijacks Docker for AI Credential Theft

CARBONATO exploits exposed Docker APIs to deploy privileged containers, persist via SSH, and use AI agent GH0ST to steal AI API keys.

cyberworldops.eu
5
SalesBleed: poisoned Salesforce Web-to-Lead forms could trigger zero-click CRM theft when Agentforce agents processed them. Patch your AI workflows. #Salesforce #CyberSecurity #PromptInjection cyberworldops.eu/en/salesble…

SalesBleed: Salesforce Lead Forms Enabled CRM Theft

SalesBleed flaws let attackers hide instructions in Salesforce Web-to-Lead forms to steal CRM data via Agentforce and post phishing in Slack.

cyberworldops.eu
10
Load more