ニュース

The Hacker News2 時間
Researchers Demonstrate How MCP Prompt Injection Can Be Used for Both Attack and Defense
Prompt injection flaws in Anthropic’s MCP and Google’s A2A protocols enable covert data exfiltration and AI manipulation.
The Hacker News4 時間
Nebulous Mantis Targets NATO-Linked Entities with Multi-Stage Malware Attacks
RomCom RAT uses bulletproof hosting and encrypted C2 for stealth; PRODAFT links it to Russian-speaking APT Nebulous Mantis.
The Hacker News7 時間
Chinese Hackers Abuse IPv6 SLAAC for AitM Attacks via Spellbinder Lateral Movement Tool
Another noteworthy tool in TheWizards' arsenal is DarkNights, which is also called DarkNimbus by Trend Micro and has been ...
The Hacker News11 時間
Meta Launches LlamaFirewall Framework to Stop AI Jailbreaks, Injections, and Insecure Code
Meta debuts LlamaFirewall with PromptGuard 2 and CyberSecEval 4 to defend AI from injection attacks and insecure code threats ...
The Hacker News8 時間
RansomHub Went Dark April 1; Affiliates Fled to Qilin, DragonForce Claimed Control
RansomHub's sudden offline status triggered affiliate migration to Qilin and cartel shifts, signaling major RaaS disruption.
The Hacker News8 時間
Customer Account Takeovers: The Multi-Billion Dollar Problem You Don't Know About
The report found a median account takeover exposure rate of 1.4% among platforms ranging from 5 million to 300 million users.
The Hacker News14 時間
Indian Court Orders Action to Block Proton Mail Over AI Deepfake Abuse Allegations
Indian court orders blocking of Proton Mail citing AI deepfakes and explicit emails, triggering national privacy concerns.