Risk
-
PhantomRaven campaign places malicious code in 126 npm packages
Researchers say a campaign codenamed PhantomRaven has placed malicious code into 126 npm packages since August 2025, using external dynamic dependencies to steal authentication tokens, CI/CD secrets and GitHub credentials; Koi Security and DCODX published analyses.
-
Python Software Foundation withdraws $1.5M NSF proposal over DEI restriction
The Python Software Foundation has withdrawn a $1.5 million NSF grant proposal after the agency attached a clause barring recipients from operating programs that “advance or promote diversity, equity, and inclusion,” a condition the PSF said conflicts with its mission.
-
Canada warns of hacktivist breaches at water, energy and farm facilities
The Canadian Centre for Cyber Security warned that hacktivists have repeatedly breached internet-exposed industrial control systems at water, oil and agricultural sites, altering control settings and prompting guidance to remove direct internet exposure, use multifactor VPNs, and report incidents through the Cyber Centre.
-
Researchers warn of ‘AI-targeted cloaking’ that can poison agentic browsers
Security researchers and hCaptcha warn of an ‘AI-targeted cloaking’ technique that serves different content to human browsers and AI crawlers, potentially poisoning models and enabling misinformation; SPLX and hTAG detail examples and risky agent behaviors.
-
UK data watchdog fines sole trader £200,000 over nearly 1m spam texts targeting people in debt
The Information Commissioner’s Office fined Bharat Singh Chand £200,000 after finding he sent 966,449 unsolicited texts about debt and energy grants from hundreds of numbers; the ICO said the messages lacked sender identification, prompted calls from a firm calling itself The Debt Relief Team, generated more than 19,000 complaints and that Chand has appealed the…
-
Researchers find 10 typosquatted npm packages delivering cross‑platform credential stealer
Researchers reported a set of 10 typosquatted npm packages, uploaded on July 4, 2025 and downloaded over 9,900 times, that deploy an obfuscated, multi-stage information stealer which harvests credentials from browsers and system keyrings on Windows, Linux and macOS.
-
Dentsu says Merkle subsidiary suffered data breach exposing staff and client information
Dentsu disclosed that U.S. subsidiary Merkle suffered a cybersecurity incident that exposed staff and client data, systems were taken offline, data were stolen and impacted individuals are being notified while an investigation continues.
-
High-severity cache-poisoning vulnerability in BIND 9; patches issued after PoC published
CVE-2025-40778 is a high-severity cache-poisoning vulnerability in BIND 9 that can allow remote attackers to inject forged DNS records. Proof-of-concept code is public and fixed versions are available; administrators are urged to apply patches immediately.
-
CISA says two Dassault DELMIA Apriso flaws are being actively exploited
CISA warned that two vulnerabilities in Dassault Systèmes’ DELMIA Apriso are being actively exploited. The flaws, CVE-2025-6205 and CVE-2025-6204, were patched by the vendor in August and have been added to CISA’s KEV catalog; U.S. federal agencies must remediate under BOD 22-01 by Nov. 18.
-
Herodotus Android malware uses human-like typing delays to evade detection
Threat Fabric has identified Herodotus, an Android malware-as-a-service that uses randomized typing delays to mimic human input and evade timing-based detection, and is being distributed via SMS to users in Italy and Brazil.
