Tag: ai
-
Black Hat news: Exposed vaults, firmware flaws, AI hacks
Check out the latest security news from the Informa TechTarget team. First seen on techtarget.com Jump to article: www.techtarget.com/searchsecurity/news/366628833/Black-Hat-news-Exposed-vaults-firmware-flaws-AI-hacks
-
Black Hat news: Exposed vaults, firmware flaws, AI hacks
Check out the latest security news from the Informa TechTarget team. First seen on techtarget.com Jump to article: www.techtarget.com/searchsecurity/news/366628833/Black-Hat-news-Exposed-vaults-firmware-flaws-AI-hacks
-
AI, Ransomware and the Security Gap for SMBs
At Black Hat, Robert Johnston and Vikram Ramesh of N-able talk about the growing security pressures on small and mid-sized businesses. They note that ransomware and credential-based attacks are climbing sharply in the mid-market. Attackers who once focused on large enterprises are increasingly targeting organizations with fewer resources, viewing them as easier entry points. In..…
-
Guess what else GPT-5 is bad at? Security
OpenAI and Microsoft have said that GPT-5 is one of their safest and secure models out of the box yet. An AI red-teamer called its performance “terrible.” First seen on cyberscoop.com Jump to article: cyberscoop.com/gpt5-openai-microsoft-security-review/
-
The Scaling Strategy that’s Failing AI Reasoning
New Architectures, Smarter Benchmarks Must Evolve Further. The era of bigger is better in AI is likely approaching its limits. Scaling reasoning AI models is proving harder than scaling perception, raising questions about architecture, context and the limits of compute. The flattening curve is a structural issue. First seen on govinfosecurity.com Jump to article: www.govinfosecurity.com/scaling-strategy-thats-failing-ai-reasoning-a-29198
-
And Now, LLMs Don’t Need Human Intervention to Plan and Executive Large, Complex Attacks
Researchers just proved LLMs can autonomously plan and execute full-scale cyberattacks, turning AI from a tool into an active threat actor. First seen on securityboulevard.com Jump to article: securityboulevard.com/2025/08/and-now-llms-dont-need-human-intervention-to-plan-and-executive-large-complex-attacks/
-
90 Prozent der deutschen Unternehmen erlebten im vergangenen Jahr Angriffe, von denen 66 Prozent erfolgreich waren
Semperis, ein Anbieter von KI-gestützter Identitätssicherheit und Cyber-Resilienz, hat seinen diesjährigen ‘Ransomware Risk Report” veröffentlicht, eine internationale Studie, an der 1.500 Unternehmen teilgenommen haben, um über ihre Erfahrungen mit Ransomware in den vergangenen zwölf Monaten zu berichten. In Deutschland erlebten 90 Prozent der befragten Unternehmen Ransomware-Angriffe, von denen 66 Prozent erfolgreich waren, was unter allen…
-
Workday research: 75% of employees will work with artificial intelligence, but not for it
Workday research finds 75% of workers like AI as a teammate, but only 30% want AI to be the boss. Trust in the technology may grow with use, but human focus, clear roles and governance are key First seen on computerweekly.com Jump to article: www.computerweekly.com/news/366629115/Workday-research-75-of-employees-will-work-with-artificial-intelligence-but-not-for-it
-
Third-Party Risk Set to Reshape AI Security
Lytical Ventures’ Taylor Margot on Autonomous Agents and New AI Defenses. As AI shifts toward autonomous agents, organizations face growing exposure from third-party systems. Strong permissioning, data orchestration and new defenses are essential to protect against opaque and potentially costly security risks, said Taylor Margot, partner at Lytical Ventures. First seen on govinfosecurity.com Jump to…
-
Russian APT28’s LameHug, a Pilot for Future AI Cyber-Attacks
While “fairly primitive”, APT28’s LameHug was a testbed for future AI-powered attacks, said two MITRE experts during Black Hat USA 2025 First seen on infosecurity-magazine.com Jump to article: www.infosecurity-magazine.com/news/mitre-russian-apt28-lamehug/
-
Black Hat NOC Expands AI Implementation Across Security Operations
Corelight’s James Pope gave Dark Reading an inside look at this year’s Black Hat Network Operations Center, detailing security challenges and rising trends, many related to increased AI use. First seen on darkreading.com Jump to article: www.darkreading.com/cybersecurity-operations/black-hat-noc-expands-ai-security-operations
-
SOCRadar Launches Agentic Threat Intelligence Platform
At Black Hat USA, which took place last week, SOCRadar launched SOCRadar Agentic Threat Intelligence. The new platform automates threat intelligence through the deployment of autonomous AI agents that proactively detect, analyse, and respond to external threats with minimal human intervention and unmatched speed and accuracy. SOCRadar Agentic Threat Intelligence is a proactive approach to…
-
SOCRadar Launches Agentic Threat Intelligence Platform
At Black Hat USA, which took place last week, SOCRadar launched SOCRadar Agentic Threat Intelligence. The new platform automates threat intelligence through the deployment of autonomous AI agents that proactively detect, analyse, and respond to external threats with minimal human intervention and unmatched speed and accuracy. SOCRadar Agentic Threat Intelligence is a proactive approach to…
-
Digitale Souveränität für Deutschland vorerst unerreichbar
Tags: access, ai, bsi, chatgpt, china, cloud, computer, cyersecurity, encryption, germany, google, governance, government, hacker, injection, Internet, risk, strategy, update, usaBSI-Präsidentin Plattner: “Wir haben technologische Abhängigkeiten an ganz vielen Stellen.” Jan WaßmuthSeine Abhängigkeit von Cloud-Lösungen, KI-Modellen und anderen Tech-Produkten aus dem Ausland wird Deutschland nach Einschätzung des Bundesamtes für Sicherheit in der Informationstechnik (BSI) so bald nicht überwinden. Da der Staat seine digitalen Systeme und Daten bis auf weiteres nicht ohne Input aus dem außereuropäischen…
-
Von der Spielerei zum Cyber-Betrug: Die neue Gefahr durch Deepfakes
Selbst technisch wenig versierte Täter können mithilfe generativer KI täuschend echte CEO-Anrufe, manipulierte Online-Meetings oder personalisierte Phishing-Kampagnen erstellen. First seen on infopoint-security.de Jump to article: www.infopoint-security.de/von-der-spielerei-zum-cyber-betrug-die-neue-gefahr-durch-deepfakes/a41663/
-
AI-Driven Vulnerability Management as a Solution for New Era
Vulnerability management (VM) is the continuous process of finding, evaluating, listing, reporting, and providing AI-driven patch recommendations for security vulnerabilities across an organization’s inventory. In practice, this means regularly scanning IT assets for known vulnerabilities, prioritizing which ones to fix first, and providing customised recommendations based on the vulnerabilities found before attackers can exploit them….…
-
9 things CISOs need know about the dark web
Tags: 2fa, access, ai, attack, automation, backup, blockchain, botnet, breach, captcha, ceo, ciso, communications, corporate, credentials, crypto, cyber, cybercrime, cybersecurity, dark-web, data, data-breach, ddos, deep-fake, defense, detection, dns, exploit, extortion, finance, fraud, group, guide, hacking, identity, incident, incident response, infrastructure, intelligence, international, jobs, law, leak, lockbit, malicious, malware, marketplace, mfa, monitoring, network, open-source, phishing, privacy, ransomware, resilience, risk, russia, saas, scam, service, strategy, tactics, technology, threat, tool, training, vpn, vulnerability, zero-dayNew groups form after major marketplaces are disrupted: International takedown efforts damage infrastructure and curb cybercrime operations by disrupting larger operations, removing major players from the ecosystem and scattering user bases.However, the dark web is highly adaptive and sophisticated actors often maintain contingency plans, including mirrors, backups, and alternative forums, according to Edward Currie, associate…
-
Poisoned telemetry can turn AIOps into AI Oops, researchers show
Sysadmins, your job is safe First seen on theregister.com Jump to article: www.theregister.com/2025/08/12/ai_models_can_be_tricked/
-
Poisoned telemetry can turn AIOps into AI Oops, researchers show
Sysadmins, your job is safe First seen on theregister.com Jump to article: www.theregister.com/2025/08/12/ai_models_can_be_tricked/
-
‘Man in the Prompt” Wenn Hacker KI-Tools heimlich manipulieren
Cyberkriminelle entwickeln mit ‘Man in the Prompt” eine neue Methode, um über Browsererweiterungen in KI-Tools einzudringen und sensible Daten zu stehlen. Besonders Unternehmen sind gefährdet, da ihre vertraulichen Informationen so unbemerkt entwendet werden können. First seen on it-daily.net Jump to article: www.it-daily.net/it-sicherheit/cybercrime/man-in-the-prompt
-
Reddit Blocks Internet Archive Amid AI Data Scraping Concerns
Reddit has announced it will restrict the Internet Archive’s Wayback Machine from accessing most of its content, citing concerns about AI companies exploiting the digital preservation service to scrape data in violation of platform policies. The move significantly limits what portions of Reddit can be archived for future reference. Major Access Restrictions Implemented The social…
-
So verwundbar sind KI-Agenten
KI-Agenten sind nützlich und gefährlich, wie aktuelle Untersuchungserkenntnisse von Sicherheitsexperten demonstrieren.Large Language Models (LLMs) werden mit immer mehr Tools und Datenquellen verbunden. Das bringt Vorteile, vergrößert aber auch die Angriffsfläche und schafft für Cyberkriminelle neue Prompt-Injection-Möglichkeiten. Das ist bekanntermaßen keine neue Angriffstechnik, erreicht aber mit Agentic AI ein völlig neues Level. Das demonstrierten Research-Spezialisten des…
-
So verwundbar sind KI-Agenten
KI-Agenten sind nützlich und gefährlich, wie aktuelle Untersuchungserkenntnisse von Sicherheitsexperten demonstrieren.Large Language Models (LLMs) werden mit immer mehr Tools und Datenquellen verbunden. Das bringt Vorteile, vergrößert aber auch die Angriffsfläche und schafft für Cyberkriminelle neue Prompt-Injection-Möglichkeiten. Das ist bekanntermaßen keine neue Angriffstechnik, erreicht aber mit Agentic AI ein völlig neues Level. Das demonstrierten Research-Spezialisten des…
-
Code-Sicherheit in Echtzeit: Checkmarx stellt Developer Assist für KI-native IDEs vor
Checkmarx, der Marktführer im Bereich autonomer, Cloud- nativer Anwendungssicherheit, stellt Checkmarx One Developer Assist vor, mit Erweiterungen für etablierte KI-native Entwicklungsumgebungen wie Windsurf by Cognition, Cursor und GitHub Copilot. First seen on ap-verlag.de Jump to article: ap-verlag.de/code-sicherheit-in-echtzeit-checkmarx-stellt-developer-assist-fuer-ki-native-ides-vor/98058/
-
Credit Karma leader shares AI governance lessons learned
Start slow and break things — that’s how the head of data and AI at the fintech says enterprises should start building AI governance frameworks. First seen on techtarget.com Jump to article: www.techtarget.com/searchitoperations/news/366628735/Credit-Karma-leader-shares-AI-governance-lessons-learned
-
Identity-first Security for AI Agents
Join us for our next industry webinar where we’ll be diving into the world of Agent AI security. First seen on securityboulevard.com Jump to article: securityboulevard.com/2025/08/identity-first-security-for-ai-agents/
-
Why AI Security Needs Continuous Red Teaming
NIST’s Apostol Vassilev Explains Need for Dynamic Response, Not Static Testing. As AI models grow in scale and power, leading to even more unpredictable outcomes, security teams are grappling with how to defend technologies that some experts can’t begin to fully comprehend. Cyber response teams are exploring the practice of continuous red teaming, said NIST’s…
-
Research shows AI agents are highly vulnerable to hijacking attacks
Experts from Zenity Labs demonstrated how attackers could exploit widely deployed AI technologies for data theft and manipulation. First seen on cybersecuritydive.com Jump to article: www.cybersecuritydive.com/news/research-shows-ai-agents-are-highly-vulnerable-to-hijacking-attacks/757319/