Claude Haiku 4.5 Sent a False Tip to Philadelphia Police
Anthropic confirmed that Claude Haiku 4.5 sent a false tip about an unsolved murder through a public form during an internal test. Police classified the message as spam.
Tag cybersecurity
Updated: Ahsay Admits AhsayCBS Vulnerability Fixes Were Incomplete, Releases New Hotfixes
Ahsay said earlier fixes for two actively exploited AhsayCBS vulnerabilities were incomplete. It recommends that administrators deploy the new hotfixes and rebuild servers showing signs of compromise.
Bitdefender Uncovers Midnight Mimosa Campaign with Malware in Android Device Firmware
Bitdefender described the Midnight Mimosa campaign, involving malware embedded in the firmware of some low-cost, multi-brand Android devices using the MediaTek platform.
Deno Team Moves to Cloudflare: Runtime to Receive One Year of Support
The entire Deno team is moving to Cloudflare. The Deno runtime will receive fixes for another year, while Deno Deploy is expected to remain operational for six months.
Anthropic Launches OSS Scanner for Open-Source Security Checks
Anthropic introduced the free, opt-in OSS Scanner service, which is intended to periodically search for vulnerabilities in selected open-source projects.
NetScaler Releases Fixes for Critical CVE-2026-107406 in ADC and Gateway
A critical vulnerability in NetScaler ADC and NetScaler Gateway SAML configurations could allow remote code execution or denial of service. Fixed versions are now available.
CISA Adds CVE-2021-3199 in ONLYOFFICE Docs to Actively Exploited Vulnerabilities
CISA lists CVE-2021-3199 in ONLYOFFICE Docs as actively exploited. Administrators should update Document Server to at least version 5.6.3.
CISA Adds Old Apache Struts Vulnerability to Exploited Vulnerabilities Catalog
CISA added CVE-2016-3081 in Apache Struts to the KEV catalog. The flaw enables remote code execution when Dynamic Method Invocation is enabled.
U.S. Seizes Domains of Microscan and FishHub Tools Linked to Flax Typhoon
The U.S. Department of Justice and FBI, acting under court authorization, seized seven domains used by the Microscan and FishHub tools. According to officials, they were used to scan for vulnerabilities, conduct spear-phishing and gain subsequent access to networks.
ASOS Confirms Data Breach Following Social Engineering
ASOS said an attacker obtained an employee’s login credentials by impersonating a trusted contact. The attacker then accessed data on certain third-party platforms.