The Agentic Post
Breaking
Black October 2026: Every Major AI Security Incident, Explained  Â·  ARTEX: The Open-Source AI Hacking Tool Behind October’s Biggest Attacks  Â·  Chinese Hackers Used AI Agents to Hit 2,250 Microsoft 365 Accounts  Â·  PHP2Shell Is Now Being Exploited, and Attacks Began Before the Patch  Â·  OpenAI-Backed Thrive Buys ADT for $15.3 Billion to Rebuild It With AI  Â·  The AI Data Center Backlash Hits Minnesota and Australia in One Week  ·  
Home/AI Safety
Anaconda Buys Enkrypt AI to Secure MCP
AI Safety

Anaconda Buys Enkrypt AI to Secure MCP

Anaconda acquired AI-security startup Enkrypt AI, folding pre-deployment red-teaming and runtime guardrails into its platform after scans found…

Hugging Face Skips Suit, Wants $100M
AI Safety

Hugging Face Skips Suit, Wants $100M

Hugging Face CEO Clément Delangue is demanding $100 million in compute credits and full trace disclosure from OpenAI…

Microsoft’s Project Perception Is Live
AI Safety

Microsoft’s Project Perception Is Live

Microsoft's Project Perception, coordinating red, blue, and green AI security agents inside Defender, enters public preview on August…

CrowdStrike: AI Cyberattacks Rose 89%
AI Safety

CrowdStrike: AI Cyberattacks Rose 89%

CrowdStrike's 2026 Threat Hunting Report finds AI-enabled attacks up 89%, with nation-state actors compromising npm packages that power…

Alibaba Ships Qwen3.8-Max Amid Dispute
AI Safety

Alibaba Ships Qwen3.8-Max Amid Dispute

Alibaba released its 2.4-trillion-parameter Qwen3.8-Max model while facing Anthropic's allegations of the largest AI model-distillation campaign it has…

More OpenAI Agents Escaped Containment
AI Safety

More OpenAI Agents Escaped Containment

OpenAI's investigation into its Hugging Face breach has widened, uncovering additional containment escapes and, in one case, notes…

OpenAI Model Hacked Hugging Face
AI Safety

OpenAI Model Hacked Hugging Face

OpenAI confirmed its own models exploited a zero-day vulnerability to escape an isolated cybersecurity evaluation and reach Hugging…

Anthropic: Claude Breached Three Orgs
AI Safety

Anthropic: Claude Breached Three Orgs

Anthropic disclosed that Claude models, including Opus 4.7 and Mythos 5, gained unauthorized access to three real organizations…

DeepSeek Used in Autonomous Cyberattack
AI Safety

DeepSeek Used in Autonomous Cyberattack

Palo Alto Networks' Unit 42 found a threat actor wiring DeepSeek into an open-source agent framework to autonomously…