Tag: ai
-
Unsloth Studio Flaw Turns Routine Model Inspection Into Code Execution
A patched Unsloth Studio vulnerability allows malicious AI models to execute arbitrary Python code during inspection, via the trust_remote_code setting. First seen on darkreading.com Jump to article: www.darkreading.com/application-security/unsloth-studio-flaw-model-inspection-code-execution
-
OpenAI apologizes for agents breaching Australian government websites without authorization
The artificial intelligence giant acknowledged it botched its response to the incidents and should have done more to promptly notify and work with the Australian government in the days after it discovered the breaches. First seen on therecord.media Jump to article: therecord.media/openai-apologizes-australia-medicare-breach
-
US is looking to weave AI into critical infrastructure for cybersecurity, national cyber director says
Sean Cairncross said CEOs need to be cognizant of how it’s being used, however. First seen on cyberscoop.com Jump to article: cyberscoop.com/national-cyber-director-ai-critical-infrastructure-cybersecurity/
-
GPT-6 Astra is More Prone to Rogue Supply-Chain Attacks
UK Agency Found GPT-6 Astra Attacked Out-of-Scope Targets in Simulated Cyber Tests. The U.K. AI Security Institute found that OpenAI’s GPT-6 Astra sometimes carried out unsanctioned supply-chain attacks in simulated environments, including against targets it had been told were out of scope. The model also created fake identities and submitted malicious code for human review.…
-
Nvidia Launches AI Agent Safety Platform to Prevent Rogue Activities
The Open Agent Safety Platform relies on both hardware and software components to monitor agent activities and quarantine unruly agents before they cause harm. First seen on darkreading.com Jump to article: www.darkreading.com/cyber-risk/nvidia-launches-ai-agent-safety-platform-prevent-rogue-activities
-
Dotted Line: How construction is dealing with cybersecurity in the age of AI
Contractors rank cybersecurity low on their priority lists. But lawyers say being prepared is critical when the inevitable breach occurs. First seen on cybersecuritydive.com Jump to article: www.cybersecuritydive.com/news/construction-cybersecurity-AI/831637/
-
How Cybersecurity Can Help Rein In Surging AI Token Costs: Experts
The same cybersecurity tools and services that are now being deployed to protect increasing AI usage may also provide the visibility needed to bring AI spending under control, solution and service provider experts tell CRN. First seen on crn.com Jump to article: www.crn.com/news/security/2026/how-cybersecurity-can-help-rein-in-surging-ai-token-costs-experts
-
Automated AI agent used to breach cybersecurity nonprofit DIVD
The Dutch Institute for Vulnerability Disclosure (DIVD) suffered an AI-driven cyberattack that the organization described as “loud and very, very messy.” First seen on bleepingcomputer.com Jump to article: www.bleepingcomputer.com/news/security/automated-ai-agent-used-to-breach-cybersecurity-nonprofit-divd/
-
Why AI won’t fix your cybersecurity problem
James Gillies, Head of Cyber Security at Logicalis UKI There is no escaping the fact that AI is creating significant opportunities for cybersecurity teams, from analysing security data and identifying suspicious activity to accelerating detection, response and remediation. However, the same capabilities are also changing the threat landscape. This comes at a time when security…
-
Amazon Bedrock AgentCore Flaws Could Expose AWS Credentials
AWS AgentCore SDK flaws could let attackers run commands in AI sandboxes and reach AWS credentials First seen on infosecurity-magazine.com Jump to article: www.infosecurity-magazine.com/news/aws-agentcore-sdk-flaws-ai/
-
OpenAI Cancels GPT-6.1 Astra Release Over Internal Safety Concerns
OpenAI has canceled the planned October release of GPT-6.1 Astra after internal testing revealed that the next-generation model did not meet the company’s safety and alignment standards. This decision underscores a growing challenge for AI developers: ensuring that highly capable autonomous systems do not exceed user intent, conceal actions, or bypass oversight. The model was…
-
RatHat’s Evolving C2 Panel Points to Malware-as-a-Service Model
RatHat’s C2 panel now builds malware and ranks victims with AI across nearly 100 deployments First seen on infosecurity-magazine.com Jump to article: www.infosecurity-magazine.com/news/rathat-c2-panel-malware-as-a/
-
OpenAI apologizes to Australia after its AI agents breached government sites
The company also detailed how some of those breaches had happened, and outlined additional measures it is taking to assess the impact of the events. First seen on techcrunch.com Jump to article: techcrunch.com/2026/09/29/openai-apologizes-to-australia-after-its-ai-agents-breached-government-sites/
-
OpenAI apologizes to Australia after its AI agents breached government sites
The company also detailed how some of those breaches had happened, and outlined additional measures it is taking to assess the impact of the events. First seen on techcrunch.com Jump to article: techcrunch.com/2026/09/29/openai-apologizes-to-australia-after-its-ai-agents-breached-government-sites/
-
OpenAI apologizes to Australia after its AI agents breached government sites
The company also detailed how some of those breaches had happened, and outlined additional measures it is taking to assess the impact of the events. First seen on techcrunch.com Jump to article: techcrunch.com/2026/09/29/openai-apologizes-to-australia-after-its-ai-agents-breached-government-sites/
-
OpenAI apologizes to Australia after its AI agents breached government sites
The company also detailed how some of those breaches had happened, and outlined additional measures it is taking to assess the impact of the events. First seen on techcrunch.com Jump to article: techcrunch.com/2026/09/29/openai-apologizes-to-australia-after-its-ai-agents-breached-government-sites/
-
Webinar: Closing the accountability gap in AI-assisted delivery
Source control records who committed code. It does not record who made the decisions behind it, and that gap is widening as AI agents take on more of the delivery process. … First seen on helpnetsecurity.com Jump to article: www.helpnetsecurity.com/2026/09/29/cleverbit-webinar-ai-assisted-delivery-accountability/
-
Meta gives small businesses an AI agent that knows their work
Meta has introduced Muse for Small Business, adding skills and connectors to its personal AI agent to help business owners get work done using the tools they already use. Muse … First seen on helpnetsecurity.com Jump to article: www.helpnetsecurity.com/2026/09/29/meta-muse-for-small-business/
-
Thales und Google Cloud erweitern Sicherheitskontrollen für agentische KI
Thales und Google Cloud erweitern ihre Zusammenarbeit zur Absicherung agentischer KI-Workflows mit Governance, Datenkontrolle und Laufzeitschutz. First seen on infopoint-security.de Jump to article: www.infopoint-security.de/thales-und-google-cloud-erweitern-sicherheitskontrollen-fuer-agentische-ki/a46564/
-
Reco raises $55M as AI agent security startups crowd the market
The round builds on a $30 million fundraise in February, taking the company’s total funding to $140 million. First seen on techcrunch.com Jump to article: techcrunch.com/2026/09/29/reco-raises-55m-as-ai-agent-security-startups-crowd-the-market/
-
Passwörter in Chatbots schaffen erhebliche Sicherheitslücken
Der neue KI-Chatbot <> von Meta fordert Nutzer dazu auf, ihre Zugangsdaten preiszugeben, um sich in Konten anzumelden und automatisierte Unterstützung bei Aufgaben zu bieten. Das Unternehmen hatte zuvor erklärt, Muse habe ‘keine Einblicke in die Passwörter oder Zahlungsmethoden der Menschen>>, und die von einem Nutzer geteilten Zugangsdaten würden ‘in einen sicheren Speicher überführt, sodass…
-
AI Agents Are Becoming Privileged Identities. Is IAM Ready?
Royal Bank of Canada, Ping Identity on Privilege, Runtime Control and AI Governance. AI agents can act independently at machine speed, creating new identity and privilege risks. Melissa Carvalho of Royal Bank of Canada and Gaurav Sharma of Ping Identity discuss how enterprises can discover shadow agents, enforce least privilege, authorize actions at runtime and…
-
OpenAI’s GPT-6 Astra ran supply chain attacks despite being told not to
OpenAI’s GPT-6 Astra carried out supply chain attacks on software outside the scope of a security test, according to the UK AI Security Institute (AISI). Anatomy of an … First seen on helpnetsecurity.com Jump to article: www.helpnetsecurity.com/2026/09/29/openai-gpt-6-astra-supply-chain-attacks-test-simulations/
-
GitHub AI Agent Uncovers 24 Android App Vulnerabilities
GitHub Security Lab has disclosed 24 vulnerabilities in Android applications discovered through its open-source AI security agent and specialized audit taskflows. The findings highlight issues that could enable covert location tracking in the OsmAnd navigation app and account takeover attacks against Wikipedia users on Android. Kevin Stubbings, a researcher at GitHub Security Lab, developed targeted…
-
New AI-Powered Botnet x47.c Steals Credentials and Drains AI Account Credits
Tags: ai, api, botnet, control, credentials, cyber, ddos, infrastructure, intelligence, malware, service, theft, threat, windowsA newly identified Windows botnet dubbed x47.c is marketing a blend of conventional DDoS tooling, credential theft, SOCKS5 proxying, fast-flux command-and-control infrastructure, and an “AI API drain” capability designed to exhaust victims’ paid artificial-intelligence service credits. Qrator Research Labs identified the previously undocumented malware platform during threat hunting. They traced its sale to an operator…
-
Proof of Concept: When AI Agents Get More Authority
Royal Bank of Canada, Ping Identity on Privilege, Runtime Control and AI Governance. AI agents can act independently at machine speed, creating new identity and privilege risks. Melissa Carvalho of Royal Bank of Canada and Gaurav Sharma of Ping Identity discuss how enterprises can discover shadow agents, enforce least privilege, authorize actions at runtime and…
-
Palo Alto Networks and NVIDIA want tighter control over AI agents
Palo Alto Networks is expanding its work with NVIDIA to help companies control what AI agents can access and do. The collaboration covers agent activity, network traffic and … First seen on helpnetsecurity.com Jump to article: www.helpnetsecurity.com/2026/09/29/palo-alto-networks-nvidia-ai-agent-security/
-
Wie viel Künstliche Intelligenz verträgt die IT-Sicherheit? – KI in der Netzwerküberwachung und beim Monitoring
First seen on security-insider.de Jump to article: www.security-insider.de/ki-in-der-netzwerkueberwachung-und-beim-monitoring-a-e25a32c79ae399c87009da309e9050e0/