Research
-
Anthropic says Chinese state-sponsored group used Claude Code AI in espionage campaign
Anthropic reported that a Chinese state-sponsored group used its Claude Code AI and a Model Context Protocol to orchestrate attempted intrusions against about 30 high-profile organizations in mid-September, succeeding in a small number of cases; Anthropic banned accounts, notified victims and said AI hallucinations limited full autonomy.
-
Over 4,300 Domains Used in Mass Phishing Campaign Targeting Hotel Guests
Researchers say a Russian-speaking threat actor registered more than 4,300 domains this year to run a large phishing campaign impersonating hotel booking services and harvesting payment data and credentials.
-
Researchers: npm registry flooded by tens of thousands of fake packages in two‑year spam campaign
Researchers have identified a two‑year spam campaign that has flooded the npm registry with tens of thousands of fake packages using a worm-like mechanism to auto-publish new packages and potentially monetize the effort via the TEA protocol; investigators say attribution is unconfirmed and registry operators have removed the packages.
-
Researchers detail Android RAT ‘Fantasy Hub’ sold as Malware‑as‑a‑Service on Telegram
Security researchers and industry trackers say an Android remote access trojan named Fantasy Hub is being sold on Russian‑language Telegram channels as a Malware‑as‑a‑Service, offering device takeover, SMS interception, APK trojanising, and subscription pricing while mirroring features seen in other Android RATs and banking trojans.
-
Amazon opens invite-only bug bounty for NOVA models to outside researchers
Amazon has launched an invite-only bug bounty program for its NOVA family of language models, allowing select researchers to test and be paid for findings on issues such as prompt injection, jailbreaking and other vulnerabilities, with the company saying the effort will help secure models integrated across Amazon and customer systems.
-
Researchers link WhatsApp-propagated Maverick malware to Brazilian banking trojans
Researchers say Maverick, a WhatsApp-propagated malware, shares code and tactics with the Brazilian banking trojan Coyote and is being spread via automated WhatsApp Web sessions, with analysts noting ties to a group called Water Saci.
-
North Korean-linked group used Google device service to wipe South Korean Android phones
South Korean researchers say the North Korean-linked KONNI group abused Google’s device-management features to remotely factory-reset Android phones, using stolen credentials harvested via phishing and RATs spread over KakaoTalk.
-
Proofpoint links new UNK_SmudgedSerpent cluster to targeted phishing of Iran experts
Proofpoint has identified a new threat cluster, UNK_SmudgedSerpent, that used political lures, impersonation and malicious installers to target academics and Iran policy experts between June and August 2025, deploying RMM tools including PDQ Connect and possibly ISL Online.
-
Researchers: Russian-linked group used Hyper-V to hide Alpine VM and bypass endpoint security
Bitdefender and Georgia CERT say Curly COMrades abused Hyper-V to run a hidden Alpine VM hosting custom implants CurlyShell and CurlCat, bypassing endpoint security and using host networking to mask malicious traffic; researchers published IoCs on GitHub.
-
Google AI agent Big Sleep credited with finding five WebKit bugs in Safari; Apple issues patches
Apple credited Google’s AI agent Big Sleep with finding five WebKit vulnerabilities affecting Safari that could cause crashes or memory corruption; Apple issued patches in iOS 26.1, macOS Tahoe 26.1, tvOS 26.1, watchOS 26.1, visionOS 26.1 and Safari 26.1 and urged users to update.
