→ Tagged

Security & Adversarial AI

13 articles

10/2026

Your Prompts Are a Record: What the OpenAI Bugging Case Shows

A federal complaint over a bugged OpenAI conference room cites chatbot prompts as evidence. What that shows about how long prompts last and who can read them.

Security & Adversarial AIAI Tools & InfrastructureIndustry Analysis

10/2026

→ Research Note 9 min

Schema-Valid Is Not Safe: What ToFu Found in 118 MCP Servers

A new preprint verified 253 bugs in MCP tools that trust whatever the model sends. Most are low-severity. The pattern matters more than the count.

Security & Adversarial AIAI Tools & Infrastructure

08/2026

→ Research Note 7 min

Deep Agents: LangChain's Open Agent Harness Layer

Deep Agents began as a Claude Code-inspired package. Its 2026 evolution shows the agent harness becoming an independently optimized layer.

AI Tools & InfrastructureAI-Assisted DevelopmentIndustry Analysis

07/2026

→ Research Note 8 min

OpenAI's Agents Hacked Hugging Face. Containment Failed

OpenAI's frontier agents breached Hugging Face during an evaluation. The lasting lesson is about containment, authority, and response.

Security & Adversarial AIAI Tools & InfrastructureIndustry Analysis

05/2026

→ Deep Dive 14 min

Anthropic Glasswing and the Gating of Superhuman Bug-Finding

Anthropic gated superhuman bug-finding to ~50 partners. The capability is real. The gate is a delay, not a fix, and the long tail pays.

Security & Adversarial AIIndustry Analysis

05/2026

→ Deep Dive 12 min

Sleeper Memory: The Prompt Injection Attack That Waits for You

LLM agents can be poisoned through documents they read today, shaping decisions made weeks later. Here's what the research actually shows.

Security & Adversarial AI

03/2026

→ Tool Review 8 min

Khaos SDK: Chaos Engineering Meets AI Agent Security Testing

Khaos SDK applies chaos engineering to AI agents — testing for prompt injection, tool misuse, and fault resilience. Here's what works and what doesn't.

Security & Adversarial AIAI Tools & Infrastructure

03/2026

→ Deep Dive 12 min

EchoLeak: Zero-Click Exfiltration Through Microsoft 365 Copilot

One email turned Microsoft 365 Copilot into a data exfiltration tool — no clicks, no user interaction. The attack bypasses every defense Microsoft built.

Security & Adversarial AIAI Tools & Infrastructure

02/2026

→ Deep Dive 11 min

DeepSeek Writes Worse Code When You Mention Tibet or Taiwan

CrowdStrike found that political trigger words increase DeepSeek-R1's vulnerability rate by 50%. The implications go far beyond one model.

Security & Adversarial AIAI-Assisted Development

01/2026

→ Experiment 13 min

LLM-Generated Passwords Are Far Weaker Than They Look

I generated passwords across seven LLMs — from Gemini 1.5 to GPT-5.4 — and measured their entropy. Centuries to crack? Try hours.

Security & Adversarial AIAI-Assisted Development

01/2026

→ Deep Dive 11 min

Clinejection: When a GitHub Issue Title Owns Your Pipeline

A GitHub issue title compromised Cline's CI/CD pipeline, stole npm tokens, and pushed malware to 4,000 devs. The first AI supply chain attack.

Security & Adversarial AIAI-Assisted Development

01/2026

→ Experiment 9 min

The Invisible Prompt: Hunting Hidden LLM Instructions on the Web

Microsoft found 50+ hidden AI instructions in commercial web pages. I built a detection pipeline, replicated the attacks, and scanned live sites.

Security & Adversarial AIIndustry Analysis

01/2026

→ Deep Dive 10 min

LLMs Hallucinate Packages. Attackers Are Registering Them.

AI coding tools invent package names that don't exist — and 43% of those names appear consistently across sessions. Attackers are registering them.

Security & Adversarial AIAI-Assisted Development