Anthropic Says Claude Hacked 3 Organizations During Cybersecurity Tests
Explorar
Noticias de IA
1057 elementos — filtrados, clasificados y sin duplicados
Anthropic says its own AI models breached three companies during security tests
Disrupting a Criminal Scam Operation
Anthropic AI Models Hacked Three Organizations During Tests
Chrome Needs Twice-a-Week Patching Thanks to AI Bug Hunting
In the Hugging Face breach, OpenAI’s hacker was noisy and fast — but not unstoppable
Security of World-Model-Based Embodied AI: A Lifecycle of Threats, Defenses, and Evaluati…
How AI is Changing Linux VPS Security for Businesses
OpenAI’s Hacking Debacle Was a Human Mistake
TriShield: Zero-Utility-Loss Defense Against Privacy Backdoors in Federated Language Mode…
AI Scammers Are Better at Building Trust Than Humans
Borrowed Strength: Best-of-N Search over a Code EncodingBreaks Self-Check Jailbreak Defen…
Guarding Organizations Against Malware Risk: A Novel Graph-Based Malware Detection Method
FakeIDet3-DB: Refining Digital Attacks and Patch Extraction for Secure ID Benchmarking
TRACE-CTI: Auditable Post-Extraction Governance of TTP Claims with Knowledge Graphs
StealthBench: Measuring Operational Stealth in Autonomous Offensive-Security Agents
Recover, Decode, Reguard: Guard-Agnostic Defense Amplification againstEncoded VLM Jailbre…
SecRespond: Benchmarking AI Agents for Real-World Post-Compromise Incident Response
Physically Real-time Infrared Attack against Optical Flow Estimation Networks
The Capability Paradox: How Smarter Auditors Make Multi-Agent Systems Less Secure
LLM Honeypot
The Hugging Face AI break-in, as told through an increasingly committed bear metaphor
The OpenAI-Hugging Face hack was worse than we thought
Creator of Test That OpenAI Models Tried to Cheat Sounds Alarm
It’s Frighteningly Easy to Jailbreak Some Frontier AI Models
AI Model Escape, Then Target Tools to Help Themselves Improve
Anthropic is finding bugs faster than Microsoft can fix them
OpenAI’s rogue AI agent didn’t stop at hacking Hugging Face
Document-borne AI worms can self-propagate through Copilot for Word
Borrowed Strength: Best-of-N Search over a Code EncodingBreaks Self-Check Jailbreak Defen…