Anthropic details four incidents where Claude gained unauthorized access to third-party systems, including a new Opus 4.6 case; METR will investigate them (Anthropic)
Covered by 3 sources
Checked for new stories 15m ago
Updates on AI Security
Every AI story we track on AI Security — 73 stories so far, each summarized in our own words and linked back to the publisher that reported it.
Pulled from 124 sources
This week
Cybersecurity2 min read
U.S. Agencies Issue Stern Rebuke of China-Based AI Companies Over Alleged Distillation
Covered by 3 sources
Business & Enterprise4 min read
Harvey Acquires Guardrails AI, Its Fourth Acquisition of 2026
Unite.AI
This month
Cybersecurity2 min read
Nvidia-Started Open Secure AI Alliance Moves to the Linux Foundation
Hacker News
Cybersecurity18 min read
PromptSonar – Execution path analyzer for AI agents and MCP servers
Hacker News
Cybersecurity5 min read
OpenAI Tells House Democrats It Is Building Automated Shutdown Capability
Covered by 2 sources
Cybersecurity4 min read
HiddenLayer nabs $100M as enterprises rush to secure their AI deployments
Covered by 2 sources
Cybersecurity5 min read
AIR raises $50M to help companies vet the skills and add-ons AI agents use
TechCrunch
Cybersecurity5 min read
A researcher hijacked Claude Code by asking it to summarise a web page
The Next Web
Cybersecurity14 min read
⚡ Weekly Recap: Chinese Spy Proxy, AI Agents Go Off-Task, Router Backdoors and More
The Hacker News
Cybersecurity7 min read
OpenAI Says Reward Hacking Drove AI Agents to Exploit Zero-Days and Breach Hugging Face
Covered by 4 sources
AI Research6 min read
The report into OpenAI’s escaping models reveals a deeper problem
Transformer News
Cybersecurity2 min read
Claude, Codex, and Hermes installed unowned code inside corporate networks
Ars Technica
Cybersecurity6 min read
Amazon Kiro Prompt Injection Can Exfiltrate Sensitive Data Through Kiro Powers
The Hacker News
AI Research6 min read
OpenAI’s training pause is convenient. That doesn't make it meaningless.
Covered by 2 sources
Cybersecurity4 min read
AWS Bedrock AgentCore enforces user context to prevent hijacked AI agents
Hacker News
AI Research3 min read
OpenAI Halts AI Training on Advanced Model as It Detects Dark Signs Emerging
Futurism
Cybersecurity5 min read
Grok exfiltrates user data when malicious instructions are encrypted
Ars Technica
Cybersecurity3 min read
Cloudflare WriteGuard Brings Fine-Grained Security Controls for MCP Servers
InfoQ (AI, ML & Data)
Cybersecurity7 min read
AI "Mind Viruses" Can Spread Between Agents Through Persistent Prompt Files
The Hacker News
Cybersecurity25 min read
Presentation: Leveraging Adversary Emulation for GenAI Red Teaming
InfoQ (AI, ML & Data)
Cybersecurity5 min read
Experts find AI agents can be tricked into 'remembering' fake facts for months — so how do we stop it?
TechRadar
Cybersecurity4 min read
Why Aren’t Any AI Companies Watching Their Frontier Models to Make Sure They Don’t Go on Hacking Sprees?
Futurism
Agents7 min read
Control agent behaviors and cost beyond a single action: new capabilities in Amazon Bedrock AgentCore
AWS Blog
AI Research7 min read
AI Recommendation Poisoning: How "Ask AI" Buttons Silently Alter LLM Memory
The Hacker News
AI Research4 min read
Nvidia is quietly staffing a new AI safety team as it doubles down on open models
Business Insider
Cybersecurity9 min read
AWS, Google, and Vercel Agent Flaws Let Attackers Trigger Tools Without Running the Model
The Hacker News
AI Research3 min read
Researchers watched OpenAI, Anthropic models take extreme measures in hacking test
Mashable
Cybersecurity4 min read
I Usually Laugh Off These AI Hacking Reports, but This One Sounds Serious and Scary
Gizmodo
Cybersecurity1 min read
Concrete Evaluations to Investigate the OpenAI Model That Hacked Hugging Face
Covered by 2 sources
AI Research24 min read
Nvidia’s Open Source Alliance Is Missing Some Key Names: OpenAI and Anthropic
Wired
AI Research1 min read
Infected Vibe-Coding: How Does an AI react to a Prompt Injection from a Different AI?
LessWrong
Cybersecurity2 min read
Cyera agrees to acquire Oasis Security for $1B to safeguard proliferating AI agents
Covered by 2 sources
Public Sector5 min read
Europe gets its AI enforcement powers on Sunday. The unit wielding them has 36 people.
The Next Web
Cybersecurity3 min read
Meta hires Assaf Keren from Qualtrics and PayPal as its new chief information security officer
The Next Web
Society & Culture3 min read
Terrified Tech Execs Are Traveling With Armed Bodyguards as AI Backlash Grows
Futurism
A fake AI agent skill passed every security scanner and reportedly reached 26,000 agents
The Next Web
Showing the 60 most recent of 73 stories on AI Security