Latest Internet & Cybersecurity News
MongoDB CVE-2025-14847 Actively Exploited Worldwide
A critical MongoDB vulnerability, CVE-2025-14847 (CVSS 8.7), enables unauthenticated attackers to leak sensitive data like passwords and API keys from server memory via malformed packets. Over 87,000 vulnerable instances exist globally, mainly in the US, China, and Europe, with 42% of cloud environments affected.
Urgent patching is recommended as exploitation is ongoing.
Cloudflare Blocks 8.3 Million DDoS Attacks in Q3 2025
Cloudflare's defenses stopped 8.3 million DDoS attacks in Q3 2025, averaging 3,780 per hour, signaling relentless threats into 2026. The surge underscores the need for robust autonomous protections against escalating volumetric assaults.
Businesses are urged to prepare for intensified DDoS campaigns.
First Confirmed Ransomware-Related Death Linked to Synnovis Attack
The Synnovis ransomware incident, confirmed in 2025, marks the first direct cybercrime death, disrupting UK healthcare services. Despite occurring in 2024, official linkage established this year highlights ransomware's lethal potential.
Previous unconfirmed cases include US Medicare patient deaths.
Cybercriminals Weaponize Preschoolers' Data in Kido International Breach
Ransomware group Radiant Group leaked personal data of 10 schoolchildren, including addresses and parents' contacts, from Kido International. This attack on preschoolers represents a new low in targeting vulnerable populations.
It exemplifies ransomware gangs' disregard for ethical boundaries.
Rise in Violence and Amputations Tied to Cybercrime in Europe
CrowdStrike reported a dramatic increase in 'violence as a service' across Europe, including amputations to coerce crypto payments from victims. Growing cryptocurrency values fuel cybercriminals' extreme tactics.
This trend signals coalescing of physical violence with digital extortion.
AI Prompt Injections Emerge as Top Security Threat
Prompt injection vulnerabilities in LLMs and AI agents like ChatGPT and Copilot allow attackers to hijack instructions, posing risks to enterprise tools. These flaws stem from inability to distinguish commands from data, affecting browsers and chatbots.
Secure communication protocols are essential to mitigate hijacking.
LLMjacking: Credential Theft for LLM Abuse Surges
Cybercriminals steal LLM API credentials for unauthorized access, racking up costs over $100,000 daily and bypassing safeguards. Microsoft sued a gang specializing in this 'LLMjacking' to build illicit services.
Victims include users of Amazon Bedrock and similar platforms.
Malware Hidden in Hugging Face AI Models Discovered
ReversingLabs uncovered malware in AI models on Hugging Face, enabling supply chain poisoning attacks on developers. Vetting sources for AI libraries is critical amid shadow AI proliferation.
Enterprises face risks from unmonitored employee AI experimentation.
Trump Administration Slashes US CISA Budget and Staff
President Trump's second term enacted deep cuts to CISA, ousting officials and threatening the CVE program. These changes rocked the US cyber establishment with global ripple effects.
Policy shifts under the new administration impacted industry widely.
UK Advances Cyber Security Bill with Ransomware Payment Ban
The UK laid its Cyber Security and Resilience Bill before Parliament in November 2025, including a ban on ransomware payments for critical sectors. It covers hospitals, schools, councils, and CNI like datacentres.
Consultations preceded the bill's progress.
Microsoft Exchange Servers Targeted in Joint Agency Alert
Australian, Canadian, and US agencies issued an emergency alert on securing Microsoft Exchange servers, a frequent attack vector. Criticism mounts over Microsoft's security obligations amid Patch Tuesday flaws.
The guide addresses historical high-impact incidents.