Updated 1h ago

generalnews.media

AI reads 36 sources every 2 hours and ranks each story by importance from 1 to 5 - so you see what actually matters.

Showing 500 most recent of 1031 stories
security

Critical Atlassian flaw lets unauthenticated attackers read files across 8 products

A critical vulnerability in Atlassian software allows unauthenticated attackers to read known files across eight products. The flaw could expose sensitive configuration or credential data, making urgent patching important. Atlassian users should update affected products immediately.

Exclusive

FBI Drops Accenture Contractor After Unpatched Flaw Led to ShinyHunters Breach

The FBI has cut ties with an Accenture contractor after a failure to apply a security patch allowed the ShinyHunters extortion group to breach systems holding sensitive data. The incident highlights how third-party vendors with privileged access remain a weak point for government agencies, even those with advanced cyber defenses. It also raises questions about vetting and patch enforcement for outside contractors handling federal data.

Exclusive

Denmark: Attackers Reached CPR Data of 8.8 Million People via Company Account

Danish authorities said attackers gained access to national CPR registry data covering 8.8 million people after compromising a company account, a figure far exceeding Denmark's roughly 6 million residents. The CPR number is Denmark's core personal identifier, used for healthcare, banking, and government services. The breach raises serious risks of identity theft and fraud at near-nationwide scale.

Exclusive

ClickFix Attack Uses Browser Cache to Bypass Windows Run Limits

A newly documented malware delivery technique called ClickFix hides malicious payloads inside the browser cache so they can be executed without triggering the Windows Run dialog's character limit. Attackers trick users into pasting commands that pull the payload from cached files, evading a common detection point. The method matters because it sidesteps a restriction many security tools relied on to flag suspicious command-line activity.

security

Microsoft Exchange Flaw Lets Authenticated Attackers Read Others' Mailboxes

A vulnerability in Microsoft Exchange allows authenticated attackers to read other users' mailboxes, potentially exposing sensitive email data. Because Exchange is widely deployed in enterprises, such access could leak confidential communications and enable further attacks within an organization.

security

Weekly Cyber Recap: NetScaler and FortiMail 0-Days, AI Coding Leaks, Ransomware Arrests

The Hacker News' weekly roundup covers actively exploited zero-days in Citrix NetScaler and Fortinet FortiMail, alongside leaks tied to AI coding tools. It also revisits the long-running Spectre v2 vulnerability and reports arrests linked to ransomware operations. The digest matters as a quick gauge of which flaws and threats defenders should prioritize.

Exclusive

Credential sprawl is growing faster than security teams can track

A Hacker News analysis warns that the number of credentials — passwords, API keys, tokens, and machine identities — is expanding faster than security teams can inventory or monitor them. The growth of cloud services, automation, and non-human identities is widening the gap between what exists and what defenders can see, leaving unmanaged secrets as a persistent attack surface.

Exclusive

Attackers exploit Realtek Jungle SDK flaw to spread Cling botnet

Security researchers report ongoing exploit attempts against a vulnerability in the Realtek Jungle SDK, a widely used software package in routers and IoT devices. The attacks install the Cling botnet, which uses STUN-based command-and-control traffic to blend in with legitimate network activity and evade detection. The campaign matters because compromised devices can be conscripted into large-scale DDoS and proxy networks.

Exclusive

AI Cracks Another Historic Cipher

A historic cipher has been broken with the help of AI, as reported on security expert Bruce Schneier's blog. It continues a trend of machine learning tools being applied to encrypted texts that resisted human cryptanalysis for years. The feat highlights how AI is increasingly capable of tasks once requiring specialized human expertise.