The Wire
Curated cybersecurity and tech news — AI-summarized, source attributed.
AI Automates Alert Triage: SOCs Shift to Engineering by 2028
Gartner predicts SOC teams will shift dramatically by 2028, with 50% more engineers than analysts—a change driven by AI agents automating traditional alert tria
AI Forces Mathematics to Reconsider What Gives Work Its Worth
The arrival of artificial intelligence in mathematics reveals a problem that predates AI itself: the field has come to measure intellectual accomplishment throu
AI Reveals First Readable Text from Vesuvius-Buried Scroll
A software engineer and classics scholar won a $50,000 prize for deciphering text from a Herculaneum scroll buried nearly 2,000 years ago. Working with micro-CT
AI Uncovers Legal System Vulnerabilities
Natural language legal codes contain exploitable vulnerabilities, much like software. Researchers in Germany used AI language models to identify numerous previo
Algorithm Over Hardware: Beating Kasparov Without Deep Blue
The article challenges the assumption that defeating Kasparov required Deep Blue's specialized hardware by demonstrating that a standard 1996 Pentium Pro runnin
Anthropic's $17B Q3 Revenue Masks Slowing Growth, $895B Infrastructure Bet
Anthropic's updated financial filing shows explosive revenue growth that's beginning to decelerate. The company pulled in roughly $17 billion in third-quarter r
Anthropic's IPO Hides Unsustainable $900 Billion in Commitments
Anthropic's November 2026 IPO filing shows strong revenue growth and a $96 billion cash position, but a closer look at the financial footnotes reveals roughly $
ARTEX AI Framework Powers WorkOS Data Breach
WorkOS, a developer platform serving 3,000+ companies including OpenAI, Anthropic, and Snowflake, suffered a breach on October 30, 2026, when a threat actor gai
Astra Weight Claims Conflate Distinct Techniques
A recent Astra paper claims the model can read raw weights to understand what networks were trained to do, generate game-playing networks that exceed PPO-traine
Attackers Hijack LLM Services to Generate $1M in Fraudulent Cloud Charges
Attackers compromise enterprise LLM infrastructure via exposed endpoints and stolen credentials, then resell access through LLM mixers—services offering subscri
Barracuda Chained Vulnerabilities Enable Root Access
Two Barracuda flaws that can be chained together create a critical attack path to full system compromise. The first vulnerability (CVE-2026-116606, CVSS 9.3) is
Barracuda Patches Critical Root Access Flaws in Email Gateways
Barracuda Networks has warned of two critical vulnerabilities in its Email Security Gateway products. The first (CVE-2026-116606) exploits an authentication wea