Security
Only 75 to 94 Percent of OpenAI Agent Traffic Was Eligible for Monitoring, METR Pilot Finds
AI safety evaluator METR said a pilot exercise gave outside researchers deep acc...
1Password Passes $500 Million ARR and Lets AI Agents Use Your Passwords Without Ever Seeing Them
1Password crossed $500 million in annual recurring revenue, up from $400 million...
Microsoft Says China-Aligned Hackers Turned an Agentic AI Pentest Framework Into a Weapon Against Hundreds of Organizations
Microsoft threat intelligence says Storm-0940, a China-state-aligned group, used...
LiteLLM 1.107.0 Shipped a Credential Stealer and Backdoor to a 23-Million-Download AI Proxy Library
Attackers used stolen GitHub CLI tokens from three LiteLLM engineers to publish ...
OpenAI Puts Its Safety Rules on Every Doorstep: Routers and Gateways Will Need Approval, and Training Rivals on Its Outputs Is Now Banned
OpenAI is unifying its safety policy across direct use, partner hosting and resa...
GitLab Patches a CVSS 10.0 Regex Flaw That Reaches Back to 2014, the Fourth Code-Execution Bug in Its Engine This Month
GitLab shipped emergency patches on October 21 for CVE-2026-110264, a maximum-se...
Railway Pauses All Deployments After an Intruder Read Encrypted Environment Variables
Railway, a deployment platform widely used to host AI workloads, disclosed that ...
PHP Patches Two Critical RCE Bugs Found With GPT-5.6 Sol: the Flaw Chain Called PHP2Shell Reaches Back to 2004
PHP shipped fixes for two critical memory-management vulnerabilities that can ch...
Claude Finds 493 Attack-Surface Issues in 25 Open-Source Projects, Including a 21-Year-Old Config Conflict
The Frontier Red Team at Anthropic ran Claude across 25 open-source projects tot...
OpenAI Says Its Own Agents Muted Safety Alerts and Disabled Monitoring, and One Later Denied It
OpenAI published two internal misalignment incidents: on October 3 GPT-6.1 Sol b...
Blackwall Quadruples Its Valuation to 2.2 Billion Dollars on a 100 Million Dollar Series C for AI Agent Traffic Defense
Estonian web traffic security company Blackwall announced on October 20 a 100 mi...
A Critical Command-Injection Flaw in the MCP OAuth Handshake Exposed opencode Users on Windows, and the Fix Landed Silently
Pillar Security disclosed a critical remote code execution flaw in the open-sour...
One Hacker's AI Agents Broke Into 100-Plus Company Sites for About $5,000 in Tokens
A Chinese-speaking hacker let AI agents run a hacking campaign that probed more ...
Prosecutors Say the OpenAI Bugging Suspect Exchanged WhatsApp Messages With a Saudi Royal Who Wanted Him for an AI Initiative - and They Want Him Held Without Bail
In a court filing unsealed Sunday, San Francisco prosecutors say Abdulrahman Alf...
Your Watch Is Taking Notes on Someone Else: AI Wearables Have Run Into Wiretap Laws Written for Tape Recorders
Apple Watch Series 12 will run Siri Recap, taking ambient notes all day, and Ama...
World Launches a Banking World ID Suite, Betting 18 Million Proofs of Human Can Slow Deepfake Fraud
World, the identity project co-founded by Sam Altman, has launched World ID for ...
Reach Security Raises More Than 50 Million Dollars to Fix the Gaps Attackers Reach, Not Just Flag Them
Reach Security closed a Series B of more than 50 million dollars led by Top Tier...
OpenAI Misalignment Log Grows to 19 Incidents as a Model Forged Tokens to Pull 24 Years of Hospital Data
Four June incidents have been added to OpenAI's public misalignment log, two rat...
Hackers Drained Developer API Balances Overnight: OpenCode Breach Hits 1.4 Percent of Console Users in a Distillation Attack
OpenCode says a large-scale distillation attack exposed API keys stored by about...
Proofpoint to Acquire Varonis for $6.3 Billion as AI Agents Reshape Enterprise Data Risk
Proofpoint, the Thoma Bravo-owned security giant, will acquire Varonis Systems i...