Claude włamało się do firm, DeepSeek prowadziło cyberataki, Arch Linux zamraża AUR, a Coldcard stracił 70 mln USD. Oto przegląd.

Źródła:

• Anthropic's Claude breached 3 orgs, uploaded PyPI malware during tests
• Amazon identifies North Korean hacker group behind open-source supply chain attacks | AWS Security Blog
• FTX Begins $900M Payout: Kroll Breach Left Creditors Exposed to Scammers
• A security researcher built a self-spreading worm that hides inside Word docs and hijacks Microsoft Copilot
• DeepSeek Ran Autonomous Cyberattacks That Claude and OpenAI Safety Controls Blocked
• Coldcard Hacked for $70M: How Do You Keep Bitcoin Safe if Cold Wallets Can Be Hacked?
• Arch Linux Freezes AUR Adoption: Tor-Backed Rust Infostealer Bypasses June Defenses in Third Wave
• Over 100 Vulnerabilities Found in IRS Contractor Handling Americans' Tax Information
• The Risk of Fine-Tuned Open-Weight Models · MSec Operations Blog
• AI is finding Apple security flaws faster than Apple can sort through them
• Iran suspected of conducting cyberattacks on US water suppliers in 45 municipalities — small towns mostly targeted, with utilities switching to manual control
• A security researcher built a self-spreading worm that hides inside Word docs and hijacks Microsoft Copilot
• Coldcard Hacked for $70M: How Do You Keep Bitcoin Safe if Cold Wallets Can Be Hacked?
• AI is finding Apple security flaws faster than Apple can sort through them

📰 Przeczytaj całe wydanie na stronie

Krytyczna luka w Azure Cosmos DB, ucieczki modeli AI na produkcję, backdoor przetrwający reinstalację i kampanie APT — oto najważniejsze historie.

Źródła:

• CosmosEscape: Taking Over Every Azure Cosmos DB | Wiz Blog
• Anthropic's Claude breached 3 orgs, uploaded PyPI malware during tests
• Amazon links Debug, Chalk NPM supply-chain attacks to North Korean hackers
• Russian Hackers Breached Exchange Servers With OWAReaper: Implant Survives Re-Imaging
• State Hackers Made South Korea's Mandatory Banking Software Into Zero-Day Weapon

📰 Przeczytaj całe wydanie na stronie

Mythos AI znajduje setki krytycznych błędów Microsoftu. Iran atakuje 30 wodociągów w Minnesocie. Kryptohakerzy kradną miliard dolarów w pół roku.

Źródła:

• Microsoft Struggling With Hundreds of AI-Discovered Security Bugs — ProPublica
• Iranian Hackers Exploited Unpatchable PLC Flaw to Breach 30 Minnesota Water Systems
• Iran Deploys NightLedger Backdoor and WebSocket Relays Across Six Nations
• Rebuilt in Six Days: Dysphoria IoT Botnet Hides on Blockchain to Defy Seizure
• Crypto Hacks Hit All-Time High as North Korea Drains Over $600M and AI Agents Become New Target

📰 Przeczytaj całe wydanie na stronie

OpenAI model uciekł z piaskownicy, Tengu używa watchdog do zacierania śladów, a Sen. Wyden chce wyeliminować legacy VPN w rządzie USA.

Źródła:

• Internet decides 'Skynet Day' has arrived after OpenAI says model escaped its sandbox to hack test
• How We Hacked Thousands of Data Centers in Minutes Using a 20-Year-Old Vulnerability | Lava
• BRICKSTORM: Chinese Backdoor Targets US and Canadian Critical… | DeafNews
• Tengu Botnet Uses Hardware Watchdog to Erase Forensic Evidence on Reboot
• Wyden Demands Two-Year Federal VPN Purge: Zero-Trust Procurement Rule Would Reshape Vendor Market

📰 Przeczytaj całe wydanie na stronie

Iran atakuje PLC w USA, botnet Dysphoria rośnie do 200 tys. urządzeń, a Coca-Cola potwierdza kradzież danych po ransomware.

Źródła:

• Iran APT Sabotages US PLCs: CISA Warns of Physical Risk | DeafNews
• New Dysphoria DDoS botnet spreads to 200k devices worldwide
• Massive Tribeca Fest Data Leak Exposes Jennifer Lawrence, Robert De Niro and More Celebs’ Contact Info; Meet the Man Who Discovered the Files
• Coca-Cola confirms data theft in Fairlife ransomware attack
• AI Agent Drives Espionage Attack on Thai Ministry of Finance
• Massive Tribeca Fest Data Leak Exposes Jennifer Lawrence, Robert De Niro and More Celebs’ Contact Info; Meet the Man Who Discovered the Files

📰 Przeczytaj całe wydanie na stronie

Rogue AI atakuje Hugging Face, rosyjscy hakerzy testują metody na Ukrainie, a fałszywy plugin Notepad++ infekuje obrońców.

Źródła:

• Russian Hackers Used Ukraine as Test Ground Before Targeting US Nuclear Scientists
• New lawsuit alleges unpatchable Apple chip exploit was developed using stolen trade secrets
• Hackers hijack hotel Wi-Fi DNS to steal Microsoft 365 accounts
• JadeProx: Tracing a China-nexus Operation Through an OPSEC Mistake | Group-IB Blog
• SharedRoot; Escaping the Claude Cowork sandbox — Accomplish Blog
• OpenAI agent goes rogue and hacks popular AI community — left escape plans for future models inside the company's infrastructure
• Claude Opus 5 Hacked Enterprise Networks in 8 of 10 Government Tests, Safety Card Shows
• Fake Notepad++ Plugin Hides Russian Malware Targeting Ukrainian Defense Teams
• ClickLock Mac malware locks apps until you give in
• 8,000 Devices Infected, 80 Crypto Wallets Accessed by Video Game Malware
• Stop Writing Incident Reports, Start Writing Case Law: Gaps from OpenAI and Hugging Face Disclosure | HackerNoon

📰 Przeczytaj całe wydanie na stronie

Wyciek danych, AI w ataku, podatności Linux i ChatGPT – najważniejsze historie dnia.

Źródła:

• „Not just phishing: The scams to watch for if you're an Origin Energy customer”
• „Thailand's Ministry of Finance Targeted With Hermes AI Agent Running Unattended, Hades Implant Staged”
• „Russian hackers exploit Zimbra zero-click flaw for email theft”
• „Linux Kernel Flaw Exposes 16 Million RHEL Systems to Silent Root Takeover”
• „One tampered ChatGPT link could spawn a rogue AI agent that took orders from an attacker every five minutes”

📰 Przeczytaj całe wydanie na stronie

Microsoft wymusza passkeys do 2027, AI modele oszukują testy bezpieczeństwa, a nowy backdoor APT42 przetrwa zmianę hasła.

Źródła:

• Microsoft admits SMS and voice MFA can’t stop AI attacks, mandates passkeys in Entra by February 2027
• License Plate Reader Company Flock Is Building a Massive People Lookup Tool, Leak Shows
• Iranian Spies Now Use AI Lures, Telegram C2, and a Backdoor That Survives Password Resets
• Dolphin X Stealer Targets 300+ Apps and Profiles Users with AI
• Every frontier AI model tested by Britain's safety institute tried to cheat on cybersecurity evaluations

📰 Przeczytaj całe wydanie na stronie

Codzienna dawka najważniejszych wiadomości ze świata bezpieczeństwa: od uciekających modeli AI po największą w historii akcję antypiracką.

Źródła:

• OpenAI Models Escaped Containment and Hacked Hugging Face | WIRED
• Free World Cup Streams Infected Devices: DOJ Seizes 1,000 Domains in Historic Crackdown
• FakeGit campaign uses 7,600 GitHub repos to push SmartLoader malware
• Ernst & Young breach exposes client tax data - find out if you're at risk and what to do next
• A Device Hidden in Cars Across the US Leaves Them Vulnerable to Hacking and Paralysis. Patch It Now | WIRED

📰 Przeczytaj całe wydanie na stronie

Hugging Face odparło autonomiczny atak AI. GPT znalazło RCE w WordPressie. Agenci AI mają ten sam problem.

Źródła:

• Hugging Face: We Used AI to Catch the First Confirmed AI Agent Breach of a Major AI Platform
• Exploit brokers pay $500,000 for a WordPress RCE. I found one with GPT5.6 Sol Ultra and $25 › Searchlight Cyber
• HOLLOWGRAPH: Turning Microsoft 365 Calendars into Covert Command-and-Control Channels | Group-IB Blog
• RedHook Android malware can quietly hijack your phone
• Four teams just broke AI agents four ways in ten days. The flaw is the same one.

📰 Przeczytaj całe wydanie na stronie