‘Zoomsday’ hack uncovered using fewer than 20 AI prompts
Explorar
Noticias de IA
1057 elementos — filtrados, clasificados y sin duplicados
Stealing Reasoning Traces from Proprietary LLM APIs
A Zoom Screen-Sharing Bug Let Anyone Take Over Other Devices on a Call
How AI is changing the vulnerability response timeline
SafeCap: Improving LVLM Safety with Image Captioning Reinforcement Learning
Defending Retrieval-Augmented Intrusion Detection Against Knowledge Poisoning and Prompt …
SALLIE: Generation-Free Hidden-State Detection of Jailbreaks and Prompt Injections Across…
TraceSafe: A Systematic Assessment of LLM Guardrails on Multi-Step Tool-Calling Trajector…
Evaluating Jailbreaking Vulnerabilities in LLMs Deployed as Assistants for Smart Grid Ope…
Compositional Threat Analysis of Latent Compromise in LLM Agent Systems: The Order 66 Sce…
Stealing Reasoning Traces from Proprietary LLM APIs
When Grammar Guides the Attack: Uncovering Control-Plane Vulnerabilities in LLMs with Str…
As AI-led attacks multiply, OpenAI launches a new cyber model
ETFs Offer 'Trusted Vehicle' for Bitcoin: Mitchnick
An OpenClaw agent reportedly hacked a gym's booking system and kicked someone off a waiti…
Over 181,000 AI meeting recordings left wide open in note taking app
Expanding Daybreak as the Cyber Defense Window Narrows
OpenAI slows down Astra development due to cybersecurity concerns
Targeted Label-Flipping and Oversampling Attacks on Federated Conditional GANs
AI Safety Fears Grow After Multiple Breaches
not much happened today
LoRAScan: Detecting Backdoor Prompts in Low-Rank Adapters for Large Language Models via D…
Genotypic Triggers: Exposing Pharmacogenomic Blind Spots via Host-Specific Backdoors in G…
Diffusion LLMs as Targets and Adversaries: Mechanistic Safety Exploits
GRM: Utility-Aware Jailbreak Attacks on Audio LLMs via Gradient-Ratio Masking
The AI safety test is becoming a safety risk
Gentoo bugzilla closed due AI bot scraper overload
OpenAI says it slowed Astra model development over security concerns
OpenAI puts the brakes on a new model because it’s supposedly too powerful
OpenAI Pauses Some Work on New Astra Model Over Cyber Concerns