RC RANDOM CHAOS

The Wire

Curated cybersecurity and tech news — AI-summarized, source attributed.

vulnerabilitysupply-chain

Keystone Destroys 18,835 Wallets After Device Verification Key Breach

An attacker breached Keystone's website through a vulnerability in a third-party monitoring tool and obtained the device verification key used in its hardware w

via The Hacker News ·
aiopen-source

Liquid's LFM3: Multimodal Models for Edge Deployment

Liquid AI released LFM3, a new generation of Liquid Foundation Models designed for multimodal edge deployment. The first release includes six 2B-parameter model

via Hacker News ·
policytech-culture

Meta Employees Condemn Electoral Bias in Brazil

Meta employees, represented through the UK's Communication Workers Union, issued a statement criticizing obvious and biased favoritism toward the Bolsonaro fami

via Hacker News ·
supply-chain

Micron Abandons 4F² DRAM, Bets on 3D Instead

Micron is discontinuing efforts to commercialize 4F² DRAM cell geometry after two decades of development, opting instead to stay with 6F² scaling. The company d

via Hacker News ·
cybersecurityidentity

Microsoft Authenticator Flaw Exposed Work Tokens to Malicious Android Apps

Microsoft patched a vulnerability in Authenticator for Android (CVE-2026-80097, CVSS 8.6) that allowed a malicious app to impersonate trusted applications and s

via The Hacker News ·
tech-culturecloud

Microsoft Extends Buyout to 60,000 Employees

Microsoft is offering voluntary buyout packages to approximately 60,000 employees—roughly a quarter of its workforce—in the US, India, Canada, and the United Ki

via Ars Technica ·
cybersecuritysupply-chain

Microsoft Removes eMudhra Certificates From Windows Trust List

Microsoft has removed seven certificates from the Windows trust list via out-of-band updates released October 26, targeting systems that do not receive automati

via BleepingComputer ·
aicybersecurity

Multiple AI Security Benchmarks Can't Replace Internal Testing

The proliferation of AI security benchmarks reflects the varied needs of different organizations, but no single metric captures all dimensions of model security

via Dark Reading ·
cybersecuritymalware

New Ransomware Gang Disrupts Major University Hospital Network

The University of New Mexico Health Sciences Center suffered a cyberattack that forced it to restrict new patient admissions, reroute ambulances, and temporaril

via BleepingComputer ·
aiprivacy

Offline AI Assistant Runs Emergency Reference on Your Computer

Survival LLM is an open-source AI assistant that operates entirely offline after initial setup. The system combines a locally-stored language model with a knowl

via Hacker News ·
cybersecurityai

OpenAI's Compaction Flaw Lets Researchers Extract Hidden Reasoning

Security firm Xint demonstrated how to extract hidden reasoning from OpenAI's reasoning models by uncovering a flaw in the company's compaction feature. OpenAI'

via The Hacker News ·
cybersecurityvulnerability

PHP2Shell Exploitation Stalls at Memory Leak Phase

Wiz researchers have been monitoring exploitation attempts targeting PHP2Shell, two newly disclosed PHP vulnerabilities (CVE-2026-109618, CVE-2026-109619). Whil

via Hacker News ·