Promptware: When AI Agents Become Attack Vectors
Samira Barnes3 weeks ago
AI security
27 stories tagged AI security.
AI. Samira Barnes3 weeks ago
Promptware: When AI Agents Become Attack Vectors
Prompt injection attacks on AI agents follow a structured kill chain — and existing legal frameworks have almost nothing to say about who's liable when it works.
AI Agents Need Decision Traces—And a Threat Model
Rachel "Rach" Kovacs2 months ago
AI. Rachel "Rach" Kovacs2 months ago
AI Agents Need Decision Traces—And a Threat Model
Neo4j's context graphs give AI agents institutional memory. That's powerful—and a threat surface. What happens when that memory gets poisoned?
Your AI Agent Knows Nothing About Your Org
Rachel "Rach" Kovacs2 months ago
AI. Rachel "Rach" Kovacs2 months ago
Your AI Agent Knows Nothing About Your Org
Context engines promise smarter AI agents—but they work by hoarding your Slack history, CTO messages, and code review patterns. Is the tradeoff worth it?
Building Secure AI Agents With Bigtable and ADK
Marcus Chen-Ramirez2 months ago
AI. Marcus Chen-Ramirez2 months ago
Building Secure AI Agents With Bigtable and ADK
Google's Bora Beran demos a healthcare AI agent built on Bigtable and ADK—and the security layers that make it worth taking seriously.
Anthropic's Claude Mythos Found Thousands of Zero-Days
Tyler Nakamura3 months ago
AI. Tyler Nakamura3 months ago
Anthropic's Claude Mythos Found Thousands of Zero-Days
Anthropic's new Claude Mythos AI discovered thousands of zero-day vulnerabilities, prompting a defensive security initiative before public release.
Nvidia's NemoClaw Bets on Engineering Basics, Not AI Hype
Rachel "Rach" Kovacs4 months ago
AI. Rachel "Rach" Kovacs4 months ago
Nvidia's NemoClaw Bets on Engineering Basics, Not AI Hype
While OpenAI and Anthropic partner with consultants to deploy AI agents, Nvidia's NemoClaw assumes developers can handle it—if we remember basic engineering.
NVIDIA's NemoClaw Promises Security, Delivers Friction
Dev Kapoor4 months ago
AI. Dev Kapoor4 months ago
NVIDIA's NemoClaw Promises Security, Delivers Friction
NVIDIA's NemoClaw wraps AI agents in strict security guardrails. But does adding enterprise-grade safety make autonomous agents too tedious to use?
The AI Agent Explosion: 35 Projects Solving Real Problems
Mike Sullivan4 months ago
AI. Mike Sullivan4 months ago
The AI Agent Explosion: 35 Projects Solving Real Problems
From security sandboxes to autonomous research pipelines, GitHub's AI agent ecosystem is addressing practical problems—not just building demos.
Every Company Needs an AI Agent Strategy Now, Says Nvidia
Zara Chen4 months ago
AI. Zara Chen4 months ago
Every Company Needs an AI Agent Strategy Now, Says Nvidia
Nvidia's Jensen Huang says every software company needs an OpenClaw strategy as Q2 becomes a race to productize AI agents for enterprise. Here's what's happening.
IBM's 2026 Threat Report: Cybersecurity Got Worse
Zara Chen5 months ago
AI. Zara Chen5 months ago
IBM's 2026 Threat Report: Cybersecurity Got Worse
IBM's latest threat intelligence index reveals alarming trends: 56% of vulnerabilities need zero authentication, ransomware groups up 49%, and AI is changing everything.
31 GitHub Projects Reveal How Developers Defend Against AI
Rachel "Rach" Kovacs5 months ago
AI. Rachel "Rach" Kovacs5 months ago
31 GitHub Projects Reveal How Developers Defend Against AI
GitHub's trending projects show developers building sandboxes, secret managers, and permission systems to control AI agents before they control everything else.
Your Company's AI Tool Might Be a Security Nightmare
Tyler Nakamura5 months ago
AI. Tyler Nakamura5 months ago
Your Company's AI Tool Might Be a Security Nightmare
AI chatbots need access to everything. Security experts Nick Selby and Sarah Wells explain why that's terrifying—and what your company should do about it.
When AI Safety Instructions Failed 37% of the Time
Bob Reynolds5 months ago
AI. Bob Reynolds5 months ago
When AI Safety Instructions Failed 37% of the Time
Anthropic tested 16 AI models with explicit safety rules. More than a third ignored them. The problem isn't the instructions—it's the assumption they'll work.
Agent Zero's New Skills Feature Makes AI Dangerously Easy
Zara Chen5 months ago
AI. Zara Chen5 months ago
Agent Zero's New Skills Feature Makes AI Dangerously Easy
Agent Zero's latest update lets anyone teach AI agents new tricks in minutes. The demo is impressive. The security warnings? Even more so.
AI Agents Are Getting God Mode—And That's a Problem
Mike Sullivan5 months ago
AI. Mike Sullivan5 months ago
AI Agents Are Getting God Mode—And That's a Problem
IBM's Grant Miller explains how AI agents with elevated permissions create security nightmares—and what actually works to prevent privilege escalation.
Google's Model Armor: AI Security Through Callbacks
Samira Barnes5 months ago
AI. Samira Barnes5 months ago
Google's Model Armor: AI Security Through Callbacks
Google's Model Armor adds security checkpoints to AI agents through ADK callbacks, intercepting threats before they reach language models.
Claude Opus 4.6 Found 500+ Critical Bugs in Open Source
Rachel "Rach" Kovacs5 months ago
AI. Rachel "Rach" Kovacs5 months ago
Claude Opus 4.6 Found 500+ Critical Bugs in Open Source
Anthropic's Claude Opus 4.6 discovered over 500 high-severity vulnerabilities in open-source code. What this means for software security going forward.
AI Compiler Writes 4,000 Commits: GitHub's Latest Tools
Samira Barnes5 months ago
AI. Samira Barnes5 months ago
AI Compiler Writes 4,000 Commits: GitHub's Latest Tools
GitHub's trending projects reveal AI-generated compilers, censorship circumvention tools, and automation systems that raise regulatory questions.
OpenAI's Codex Launch Feels Like Playing Catch-Up
Marcus Chen-Ramirez5 months ago
AI. Marcus Chen-Ramirez5 months ago
OpenAI's Codex Launch Feels Like Playing Catch-Up
OpenAI released Codex, its coding agent app. Industry experts aren't impressed—it's table stakes, not innovation. Plus: AI agents got a Reddit, and it went badly.
OpenClaw Gives AI Agents Root Access to Your Machine
Dev Kapoor5 months ago
AI. Dev Kapoor5 months ago
OpenClaw Gives AI Agents Root Access to Your Machine
OpenClaw lets you run autonomous AI agents with full system access. The security implications are fascinating—and the project handles them honestly.