Technology ❯ Artificial Intelligence ❯ AI Security

Vulnerabilities

Zero-Click Attacks Privacy Risks Development Tools Browser Agents Prompt Injection Testing Bypassing Restrictions Hacking Microsoft Copilot Exploitation Techniques Cryptographic Context Injection ExploitGym Risk Management Data Leakage

4 ARTICLES

last mo.

Encrypted Payloads Let Web Pages Make Grok Leak User Session Data

The demonstration shows text-only safety checks miss instructions recovered by a model’s runtime, signaling a need for platform-level egress controls, provenance separation, and confirmed tool-argument gating.

5 ARTICLES

last mo.

LiteLLM Supply‑Chain Breach May Have Exposed Thousands After 40‑Minute Malicious PyPI Window

6 ARTICLES

4mo ago

Pwn2Own Berlin Uncovers 47 Zero‑Days, Awards $1.3 Million

8 ARTICLES

7mo ago

OpenAI to Acquire Promptfoo to Build Automated Security Into Frontier Agents

9 ARTICLES

8mo ago

Microsoft Fixes 'Reprompt' Flaw That Enabled One-Click Data Theft From Copilot Personal

9 ARTICLES

last yr.

Anthropic Adds Native File Creation to Claude in Preview Rollout

5 ARTICLES

last yr.

Microsoft Fixes Zero-Click EchoLeak Flaw in Copilot as Researchers Call for AI Agent Redesign