Tag: artificial intelligence
1911 articles
Evaluating AI in Security Operations Requires New Framework
When evaluating AI in security operations, it's crucial to determine if it can deliver accurate verdicts across various scenarios and attack surfaces - and surprisingly, verdict quality only improves dramatically once a certain threshold of relevant data, such as identity and context, is reached. Below that threshold, no amount of fine-tuning can compensate.
Hugging Face Breach Exposes AI Agent's Role in Autonomous Attack
In a chilling breach, Hugging Face revealed that an autonomous AI agent was behind a sophisticated attack that began with a simple malicious dataset upload, exploiting vulnerabilities to execute code and launch a swarm of actions across short-lived sandboxes. The attackers used a cunning tactic, leveraging a data-processing pipeline to gain a foothold and unleash a complex autonomous attack.
China's AI Governance Push Sparks Global Concerns
President Xi Jinping is calling for a people-centered approach to AI governance, urging global cooperation to harness its power for the greater good and ensure a just and equitable future for all. He envisions a collaborative framework that promotes openness, inclusiveness, and human control to drive shared prosperity and common security.
Hugging Face Breach Exposes AI Model Risks
Hugging Face revealed a shocking breach that highlights the hidden dangers of AI models, admitting to unauthorized access to internal datasets and credentials used by its services. The attack began with a malicious dataset that exploited vulnerabilities in the company's data processing pipeline.
Pentagon Networks Face AI-Driven Threats
The Pentagon faces a daunting cyber threat landscape, with dozens of interconnected networks across the services vulnerable to AI-enabled attacks. This complex web of connections multiplies the risk, making it a pressing challenge for cyber planners to defend against increasingly sophisticated threats.
Australia Wrestles with Sovereign AI Capability Plan
Australia is taking a bold step towards tech independence, with Prime Minister Anthony Albanese vowing to turn the country into a hub for large-scale AI training and cement its sovereignty in the field. By hosting cutting-edge AI data centres, Australia aims to break free from reliance on foreign tech and secure its economic resilience.
Military Autonomy Spurs Race for Trusted Information Infrastructure
The future of military capability hinges on a robust trusted information infrastructure that seamlessly connects autonomous systems, enabling secure and efficient data exchange. This backbone of innovation is now driving Western defense plans and funding.
GRC AI Tools Fall Short on Enterprise Readiness
Most organizations are flying blind when it comes to GRC AI tools, with a staggering 87% of IT and security pros admitting they can't fully see the AI tools active within their organization. This visibility gap poses a foundational risk, making it harder to rely on other controls and assurances.
South Korea Develops Sovereign AI Model for Bug Detection
South Korea is racing to develop its own AI model for bug detection by 2026, driven by concerns that relying on foreign models could leave the country vulnerable at critical moments. The homegrown AI model aims to rival leading security-focused models, ensuring the nation has control over its bug-finding capabilities.
OpenAI GPT-5.6 Model Exposes Users to Unintended File Deletion Risk
Beware of the GPT-5.6 model's alarming bug that can wipe out your files in an instant, as shocking incidents from tech investor Matt Shumer and software engineer Bruno Lemos reveal. Users are reporting devastating file deletions, with one victim losing almost all of their Mac's files and another having their entire production database erased.
AI Models Vulnerable to Poisoning for Under $100
A cybersecurity expert recently discovered that AI models can be easily manipulated to behave maliciously, with a backdoor installable in just an hour for under $100. This startling vulnerability was uncovered through a simple fine-tuning test that quickly escalated into a full-blown security threat.
Rethinking Privacy Rules in AI Era
As AI continues to shape our world, a crucial question emerges: will our privacy protections rely on individuals controlling their data or on companies being held accountable for how they collect and use it? By shifting the focus from personal control to corporate responsibility, we can create a more effective and fair approach to safeguarding our personal information.
Security Teams Must Adapt as AI Agents Disrupt Traditional Playbook
The era of predictable enterprise security is over: AI agents are autonomously accessing production data, forcing security teams to rethink everything they thought they knew. With AI agents blurring the lines between sanctioned and unsanctioned activity, traditional security playbooks are no longer effective.
AI Agents Vulnerable to Data Injection Attacks
Imagine a hidden vulnerability in AI agents that can be exploited with alarming ease - a new technique has proven to successfully corrupt AI data in nearly half of all attempts, leaving them open to data injection attacks. Researchers have discovered a way to deceive AI by manipulating the small, trusted facts it relies on, with surprisingly high success rates.
OpenAI Bolsters GPT-5.6 with Automated Red-Teaming Model
OpenAI just unveiled GPT-Red, an automated red-teaming model that's a game-changer in detecting prompt injection attacks, helping to shield its GPT models from vulnerabilities. By mimicking human red-teaming tactics, GPT-Red identifies and feeds back crucial insights to strengthen model defenses before they go live.
Harnesses Unlock AI's Cybersecurity Potential
Imagine a cybersecurity system that can launch complex attacks autonomously, accelerating threat detection and response - a recent experiment by Cato Networks has made this a daunting reality. By harnessing AI's potential, their production tool paired with OpenAI models produced end-to-end attack chains in as little as 40 minutes.
Agencies Modernize Identity Infrastructure to Counter Emerging Threats
Federal agencies are racing against the clock to modernize their identity infrastructure, securing legacy systems while navigating the rapid evolution of AI and emerging post-quantum threats. As AI continues to advance at breakneck speed, agencies must build adaptable cybersecurity strategies to stay ahead of the threat landscape.
Microsoft Disrupts 570 Security Flaws in Record Patch Tuesday Release
Microsoft just dropped a record-breaking Patch Tuesday update, fixing a whopping 570 security flaws - nearly triple the number from last month - with nearly 60 of them rated critical, allowing attackers to take remote control of your Windows device with ease. This massive update, powered by AI-driven vulnerability discoveries, also patched three zero-day vulnerabilities already being exploited in the wild.
Marketing AI Adoption Surges, Governance Gaps Persist
The AI marketing revolution is in full swing, with 70% of companies now deploying custom AI agents to tackle real marketing tasks - but as adoption surges, a new challenge emerges: how to scale and govern these powerful tools effectively. With only 3% of businesses not using marketing AI at all, the focus has shifted from "should we adopt?" to "how do we manage?"
Pentagon Targets Edge AI with New Battlefield Data Strategy
Bala Selvam shook up the Special Operations Command Pacific by ditching vague AI chatter and getting commanders to zero in on exactly what they needed to achieve. He forced them to prioritize, boiling down dozens of requests into a concise list of core requirements.
US Unveils 'Gold Eagle' AI Cyber Threat Clearinghouse
The US Treasury Department has launched Gold Eagle, a cutting-edge AI cyber threat clearinghouse that brings together government and industry to share vital threat information and safeguard America's financial institutions. This innovative hub, created through a White House executive order, aims to close vulnerabilities and protect the integrity of the US financial system.
DevSecOps Becomes Essential for Modern Government IT
As AI-assisted development accelerates delivery timelines, it's also introducing new risks, with vulnerability disclosures related to AI-assisted development skyrocketing in 2026 and leaving security leaders scrambling to harden defenses. With adversaries using AI to fuel attacks, the need for DevSecOps has never been more pressing.
Pentera Injects Validation into AI-Driven Security Workflows
Pentera is revolutionizing AI-driven security by injecting validation into workflows, empowering teams to turn disconnected risk signals into decisive action against real attack paths. By safely emulating attacker techniques, Pentera provides the evidence needed to transform guesswork into effective security measures.
AI Fuels End-to-End Cyberattacks With Expanded Role Across Intrusion Stages
Criminal groups are now using AI as the main driver behind massive cyberattacks, breaching government agencies and carrying out thousands of commands with minimal human oversight. This marks a significant shift from AI as a supporting tool to a primary operator in end-to-end cyberattacks.