AI Security · Page 7
deepfakes, AI-powered scams, model abuse, and keeping AI systems and their users safe
Apple claims ex-employee used stolen circuit designs at OpenAI, then tried to destroy the evidence
A recovered work laptop appears to show the engineer used confidential Apple files inside OpenAI. Apple is now asking a court to halt OpenAI's hardware work while the case plays out.
Your AI Agent Passed the Security Check. That Is Not Enough.
Logging in is the easy part. The real danger starts the moment an AI agent begins acting on its own, and most enterprise security tools are not watching.
OpenClaw 2.0 es más fácil de instalar. No es mucho más seguro.
La popular herramienta de agente de IA de código abierto recibe su mayor actualización, con una interfaz más elegante y configuración más simple. Los problemas de seguridad persisten en gran medida.
AI Origin Labels Can Mislead: Why Your Model's Roots Matter
New research reveals AI models can carry hidden risks from parent models. How should organisations respond?
AI Companies Warn Cyberattacks Could Overwhelm Defences Within Months
OpenAI, Anthropic and over a hundred companies have signed a joint letter saying organisations have very little time to prepare for AI-assisted hacking. Federal officials confirm hackers hit more than 100 water systems in July.
Los incidentes de "pérdida de control" de IA casi se duplicaron en julio, superando los 300 casos
Un proyecto de monitoreo que registra informes reales de mal funcionamiento de IA reporta que los incidentes de engaño, instrucciones ignoradas y búsqueda de objetivos dañinos se dispararon bruscamente el mes pasado, y la gravedad está empeorando.
AI Model Bypasses Gym Booking Limit in Tests
Claude Opus 4.6, an AI model from Anthropic, bypassed an online booking cap in controlled tests, highlighting security concerns.
AI Steps In as Vulnerability Management Struggles with Volume
Software flaws are multiplying faster than security teams can handle them. AI is being sold as the fix. Here's what the pitch gets right and what it glosses over.
The AI attack that tops every expert threat list barely shows up in real incident data. Here is why that gap matters.
Two security researchers compared expert opinion against 6,639 real-world AI security incidents and found the rankings barely agree. The most consequential finding: the most dangerous attack leaves no trace a scanner can find.
Lawsuit accuses xAI of training Grok on child sexual abuse images
A survivor filed a federal complaint this week alleging that xAI, the company behind the Grok chatbot, used images of her abuse to train its AI models. The case puts a spotlight on where AI companies source their training data and what safeguards, if any, they apply.
Over 100 AI Companies Sign Letter Warning of AI-Powered Cyber Attacks on Hospitals and Infrastructure
OpenAI, Anthropic, Google and Microsoft are among more than a hundred signatories saying AI-enabled attacks are accelerating, and that neither business nor government can handle them alone.
AI Agents Have Now Gone Rogue and Hacked Real Companies at Least 17 Times
What started as a one-off lab accident has become a pattern. Here is every case where an AI, left to complete a task, broke out and attacked systems it was never meant to touch.
Agentes de codificación de IA instalaron código desconocido dentro de redes corporativas a través de una falla en un estándar web poco conocido
Investigadores tendieron una trampa en archivos que los agentes de IA leen automáticamente. Docenas de empresas, algunas de ellas de Fortune 500, cayeron directamente en ella.
OpenAI Bans Russian Accounts That Used ChatGPT to Build a Fake Academic Institution
A covert influence operation used OpenAI's chatbot to write social media posts, copy real research papers, and construct a phoney think-tank designed to make pro-Russia narratives look credible.
Equifax Gets 20 Million Security Alerts a Day. Here Is How AI Handles Half of Them.
The credit bureau is using AI to sort, triage and close security incidents automatically, and cut the time analysts spend on each case by 61%.