How 1Password and NVIDIA Keep Secrets From Agents
1Password brokers agent logins so the agent never holds the password. NVIDIA's OpenShell does the same for self-hosted agents. Where both still fall short.
Tag
#ai-agents
Meta Muse's Pre-Launch VM Escape: An Agent Risk Class
404 Media reports, citing one unnamed source and internal posts, that Meta rushed to fix Muse VM escape bugs before launch. What it means for agents.
AI Agents in Your Text Messages: The New Privacy Surface
A wave of SMS-resident AI agents read your calendar, email, and contacts. TechCrunch's October 2026 survey maps more than 15 vendors and what they can reach.
Apple Tightens macOS Full Disk Access Over AI Agents
Apple will add new controls to macOS Full Disk Access, citing AI agents, weeks after a columnist said Meta's Muse read his messages. Meta disputed it.
FTC's First AI Agent Probe Names OpenAI, Anthropic
On Sept 30, 2026, the FTC opened a probe into OpenAI, Anthropic and other AI labs, reported as the first US enforcement action on rogue AI agents.
Inside Meta Muse's Privacy Architecture
Meta's personal AI agent runs in a dedicated VM with a Sentinel guard. The promise of ad-system isolation comes from a company with $23B in recent fines.
An Agent Hacked a Gym API. Labs Are Gating Cyber Models.
In one week, an OpenClaw agent canceled a stranger's gym booking via a missing auth check, and OpenAI moved GPT-5.6-Cyber behind a partner-only Red tier.
Claude Cowork's SharedRoot escape: a one-prompt path to your Mac
Accomplish's Oren Yomtov chained CVE-2026-46331 to escape Claude Cowork's macOS sandbox in a single short message and reach the host filesystem.
Claude's web_fetch Tool Leaked User Data Through a Honeypot
Anthropic's web_fetch tool let a prompt-injection honeypot walk Claude through user profile URLs and pull a user's name, city, and employer.
GitHub's AI Coding Agent Leaked Private Repos With a Single Issue
Noma Labs' GitLost write-up shows a single public-repo issue can coerce GitHub's coding agent into leaking private repo contents.
Alibaba's Elements Claw Discovers Four New Superconductors
A 1B-parameter foundation model trained on 125M crystal structures screens 2.4M compounds and laboratory-confirms four new superconductors.
The 'AI Coworker' Trap: Managers Catch 18% Fewer Errors
Calling an AI an 'AI employee' made 1,261 managers miss 18% more errors and push 44% more questionable work upward. The 'coworker' framing is a safety issue.
An AI Agent Deleted a Production Database in 9 Seconds
A Cursor agent running Claude Opus found an overprivileged API token, guessed wrong, and wiped a company's data and backups. The real failure wasn't the model.
AI Security Roundup: OpenClaw's Nine CVEs in Four Days
OpenClaw collected nine CVEs in four days with 135,000 instances exposed. Plus: GitHub RCE, Flowise exploitation, and CrewAI trust failures.
OpenCode Has 207K GitHub Stars and Costs Nothing
The open-source AI coding agent lets you bring any model to the terminal. We break down what works, what doesn't, and who should use it.
The FDA Just Created a New Category of Violation: AI Over-Reliance
A drug manufacturer told federal inspectors the AI never told them about a basic legal requirement. The FDA was not amused.
Meta Is Recording Every Keystroke Its Employees Make to Train AI Agents
Meta's Model Capability Initiative captures mouse movements, keystrokes, and screenshots from employee computers. The goal: build AI agents that can replace the workers generating the training data.
AI Prediction Scorecard: November 2025 Edition — Billion Agents and White-Collar Bloodbaths
We tracked the boldest AI predictions from November-December 2025 and scored them against April 2026 reality. The agents didn't show up. The jobs did disappear.
AI Security Roundup: Azure AI Foundry Scores a Perfect 10, Langflow Hijacked in 20 Hours, LiteLLM's Backdoor Cleanup
Microsoft's Azure AI Foundry hit with a maximum-severity privilege escalation, Langflow exploited within hours of disclosure, and LiteLLM discloses three vulnerabilities after surviving a supply chain attack.
OpenClaw's Security Meltdown: 9 CVEs in 4 Days, 135K Exposed Instances, and a Poisoned Marketplace
The fastest-growing GitHub project ever just became the biggest AI agent security disaster of 2026. Here's what happened and why it matters.
AI Prediction Scorecard: October 2025 Edition — Six Months of Receipts
We tracked the boldest AI predictions from October 2025 and scored them against April 2026 reality. Spoiler: the crystal balls are still broken.
9 CVEs in 4 Days: OpenClaw's Security Crisis Deepens
OpenClaw went from one CVE to nine in four days, with 12% of its marketplace confirmed malicious. Plus: ChatGPT's patched DNS exfiltration flaw.
OpenClaw: The AI Agent That Broke Every Record and Then Broke Security
The fastest-growing open source project in GitHub history has become 2026's first major AI security disaster, with 135,000+ exposed instances, 9 CVEs in 4 days, and malware-laced skills.
OpenClaw's March Meltdown: 9 CVEs in 4 Days, 12% of Marketplace Skills Are Malware
OpenClaw's security crisis escalates with nine new vulnerabilities including a CVSS 9.9 admin bypass, plus researchers confirm nearly 1 in 8 marketplace skills steal user data.