Tag: ai
-
Bei internen KI-Tests: OpenAI hat versehentlich Hugging Face gehackt
Der kürzlich bekannt gewordene Cyberangriff auf Hugging Face kam von OpenAI. Das Unternehmen hatte seine KI-Modelle bei internen Cybertests nicht im Griff. First seen on golem.de Jump to article: www.golem.de/news/bei-internen-ki-tests-openai-hat-versehentlich-hugging-face-gehackt-2607-211137.html
-
Bei internen KI-Tests: OpenAI hat versehentlich Hugging Face gehackt
Der kürzlich bekannt gewordene Cyberangriff auf Hugging Face kam von OpenAI. Das Unternehmen hatte seine KI-Modelle bei internen Cybertests nicht im Griff. First seen on golem.de Jump to article: www.golem.de/news/bei-internen-ki-tests-openai-hat-versehentlich-hugging-face-gehackt-2607-211137.html
-
Bei internen KI-Tests: OpenAI hat versehentlich Hugging Face gehackt
Der kürzlich bekannt gewordene Cyberangriff auf Hugging Face kam von OpenAI. Das Unternehmen hatte seine KI-Modelle bei internen Cybertests nicht im Griff. First seen on golem.de Jump to article: www.golem.de/news/bei-internen-ki-tests-openai-hat-versehentlich-hugging-face-gehackt-2607-211137.html
-
Ransomware JadePuffer greift KI-Modell-Daten mit EncForge an
Der autonome KI-Agent JadePuffer nutzt die Erpressersoftware EncForge zur Verschlüsselung von Trainingsdaten und Vektordatenbanken. Sysdig schlägt Alarm. First seen on it-daily.net Jump to article: www.it-daily.net/it-sicherheit/cybercrime/ransomware-jadepuffer-ki
-
Für KI reißen Unternehmen etablierte Sicherheitsmauern ein
Management Summary Der breite Einsatz von künstlicher Intelligenz verändert die Sicherheitsarchitektur von Unternehmen grundlegend: KI-Systeme benötigen umfangreichen, schnellen Datenzugriff und können dadurch etablierte Schutzmechanismen wie rollenbasierte Zugriffssteuerung, Row-Level Security sowie kontrollierte Datenfreigaben unterlaufen. Damit verschiebt sich die zentrale Governance-Frage von »Wer darf welche Daten sehen?« zu »Was darf die KI aus welchen Daten ableiten ……
-
Zwei Drittel der Unternehmen verschieben die Einführung von Microsoft Copilot wegen Datenschutzrisiken
Zwei von drei (66 %) Unternehmen haben Bedenken, dass Microsoft Copilot vertrauliche Daten offenlegen könnte und haben deshalb die Einführung des KI-Assistenten verschoben oder komplett gestrichen. Zudem befürchten nahezu drei Viertel (73 %) der Benutzer, dass KI bereits intern sensible Informationen preisgibt. Zu diesen und weiteren Ergebnissen kommt der neue Report »State of Microsoft… First…
-
Google Unveils Gemini 3.5 Flash Cyber to Find and Fix Software Vulnerabilities Faster
Google has introduced Gemini 3.5 Flash Cyber, a lightweight AI model designed to improve cybersecurity by helping defenders identify, validate, and patch software vulnerabilities more efficiently. Built on Gemini 3.5 Flash and optimized for security tasks, Flash Cyber aims to deliver a cost-effective alternative to larger AI models while supporting large-scale vulnerability analysis. First seen on thecyberexpress.com Jump to…
-
Small teams are the heaviest users of AI coding agents
Tags: aiThe pull request arrives with the tests already run and the description already written, the work of an agent that handled the whole thing on its own. Somebody still has to … First seen on helpnetsecurity.com Jump to article: www.helpnetsecurity.com/2026/07/22/users-of-ai-coding-agents/
-
AccuKnox Wins Best AI Startup Award for Enterprise Agentic AI Security at BSides Bangalore
Bangalore, India, July 22nd, 2026, CyberNewswire AccuKnox announced it has won the number one Startup Award at Security BSides Bangalore 2026, marking the second consecutive year the company has topped the category after also winning in 2025. The back-to-back recognition affirms AccuKnox’s standing among the region’s leading cybersecurity startups. AI Security Adoption For AccuKnox, this award…
-
Yubico Launches YubiKey 5.8 With Hardware-Backed Authorization for AI Agent Workflows
Yubico has released the YubiKey firmware version 5.85.85.8, expanding its hardware security key platform beyond phishing-resistant authentication. This update introduces verifiable, hardware-backed authorization for digital signatures, identity wallets, payment confirmations, and AI agent approval workflows. Announced on July 21, 2026, this firmware update aims to help enterprises verify not only who accesses an application but…
-
Microsoft Azure DevOps MCP Flaw Lets Hidden PR Comments Hijack AI Review Agents
A single invisible comment in an Azure DevOps pull request can turn a reviewer’s own AI coding agent against them, driving it into projects the attacker has no rights to reach and quietly leaking what it finds.The flaw is in Microsoft’s official Azure DevOps MCP server, and it works because one of its tools returns…
-
OpenAI Says Its AI Models Escaped Sandbox, Targeted Hugging Face to Cheat Benchmark
OpenAI on Tuesday said a combination of its artificial intelligence (AI) models, including GPT-5.6 Sol and an “even more capable pre-release model,” was behind the security incident that targeted Hugging Face’s production infrastructure last week.The AI company said the models were operating with “reduced cyber refusals for evaluation purposes” that might otherwise limit their ability…
-
Cyberangriffe auf Kanzleien – KI am Telefon, Malware im Anhang
First seen on security-insider.de Jump to article: www.security-insider.de/cyberangriffe-kanzleien-ki-anrufe-malware-a-58ac369bc714d172310afb86c9f43b15/
-
OpenAI says its AI models hacked Hugging Face during testing
OpenAI says its AI models, including GPT”‘5.6 Sol and a pre-release model, hacked into the Hugging Face artificial intelligence repository while being tested in a sandboxed testing environment. First seen on bleepingcomputer.com Jump to article: www.bleepingcomputer.com/news/security/openai-says-its-ai-models-hacked-hugging-face-during-testing/
-
Google Launches Gemini 3.5 Flash Cyber to Find, Validate, and Patch Critical Vulnerabilities
Google has introduced Gemini 3.5 Flash Cyber, a lightweight AI model specifically designed to help security teams discover, validate, and patch critical software vulnerabilities at scale. Announced on July 21, 2026, this model builds on Gemini 3.5 Flash and is optimized for security workflows. It enables agents to inspect large codebases, explore numerous execution paths,…
-
AccuKnox Wins Best AI Startup Award for Enterprise Agentic AI Security at BSides Bangalore
Bangalore, India, 22nd July 2026, CyberNewswire First seen on hackread.com Jump to article: hackread.com/accuknox-wins-best-ai-startup-award-for-enterprise-agentic-ai-security-at-bsides-bangalore/
-
FakeGit verbreitet SmartLoader über Tausende GitHub-Projekte
In der Kampagne FakeGit verbreiten gefälschte GitHub-Repositorys und KI-Server die Schadsoftware SmartLoader. Auch KI-Agenten fallen auf den Trick herein. First seen on it-daily.net Jump to article: www.it-daily.net/it-sicherheit/cybercrime/github-projekte-fakegit
-
AI can’t fix cybersecurity’s hiring problem
Organizations are redefining cybersecurity roles through workforce frameworks and placing greater emphasis on verified skills as AI and new regulatory requirements change … First seen on helpnetsecurity.com Jump to article: www.helpnetsecurity.com/2026/07/22/cybersecurity-workforce-trends-report/
-
Cloud operations become the next big role for agentic AI
Companies are using agentic AI to manage growing application environments, automate routine tasks, and support decisions. Business and IT leaders increasingly see the … First seen on helpnetsecurity.com Jump to article: www.helpnetsecurity.com/2026/07/22/agentic-ai-cloud-operations-report/
-
OpenAI Exploits Zero-Day to Gain Internet Access and Compromise Hugging Face Servers
OpenAI has revealed that during an internal evaluation of advanced cyber capabilities, AI agents exploited a zero-day vulnerability, escaped a constrained research environment, and compromised parts of Hugging Face’s production infrastructure. While Hugging Face detected and contained the activity, OpenAI’s internal security team also identified unusual behavior during the assessment. OpenAI Compromise Hugging Face Servers…
-
“Stop asking whether AI works and start asking what it can reach”: C1 CISO on AI agent security
First seen on scworld.com Jump to article: www.scworld.com/news/stop-asking-whether-ai-works-and-start-asking-what-it-can-reach-c1-ciso-on-ai-agent-security
-
Mithra AI launches to verify data before AI models produce answers
First seen on scworld.com Jump to article: www.scworld.com/brief/mithra-ai-launches-to-verify-data-before-ai-models-produce-answers
-
AI is making cloud advisory predictive
First seen on scworld.com Jump to article: www.scworld.com/perspective/ai-is-making-cloud-advisory-predictive
-
7 mistakes companies make deploying AI agents
Tags: aiFirst seen on scworld.com Jump to article: www.scworld.com/perspective/7-mistakes-companies-make-deploying-ai-agents
-
Bridging the divide: The convergence of AI and infosec
First seen on scworld.com Jump to article: www.scworld.com/resource/bridging-the-divide-the-convergence-of-ai-and-infosec
-
Guide helps organizations evaluate AI in security operations centers
First seen on scworld.com Jump to article: www.scworld.com/brief/guide-helps-organizations-evaluate-ai-in-security-operations-centers
-
Hugging Face uses GLM 5.2 to investigate AI agent-driven cyberattack
First seen on scworld.com Jump to article: www.scworld.com/news/hugging-face-uses-glm-5-2-to-investigate-ai-agent-driven-cyberattack
-
Head of federal AI testing lab resigns after three months
Tags: aiFirst seen on scworld.com Jump to article: www.scworld.com/brief/head-of-federal-ai-testing-lab-resigns-after-three-months
-
Cyber resilience fuels AI growth, but CISOs see key gaps
First seen on scworld.com Jump to article: www.scworld.com/native/cyber-resilience-fuels-ai-growth-but-cisos-see-key-gaps
-
Critical ServiceNow AI flaw exploited days after patch release
First seen on scworld.com Jump to article: www.scworld.com/news/critical-servicenow-ai-flaw-exploited-days-after-patch-release