Advertisement
AI Agent Social Engineering: The Future of BEC by 2026
Future BEC attacks may target AI agents with growing business system authority, necessitating new defensive strategies for 2026.
Securing AI Agent Permissions: Preventing Cloud Data Loss
AI agents can escalate privileges and perform unauthorized actions, leading to data loss. Learn how to implement granular controls for tool calls.
Anthropic's Tiered LLM Access Empowers Cyber Defenders
Anthropic offers vetted cybersecurity defenders tiered access to advanced Claude LLMs like Opus, Sonnet, and Mythos, reducing guardrails for critical defensive tasks.
Addressing Misconceptions: AI Agents and Security Risk Attribution
Analysis debunks 'rogue AI' claims, urging security professionals to treat AI agents as untrusted software for effective risk management.
doxx.net Raises $38M to Secure AI Agents on the Internet
doxx.net secures $38M to launch Agentic Defined Networking (ADN), a private platform protecting AI agents from malicious destinations and ensuring user privacy.
OpenAI Security Incident Highlights Transparency Gaps
OpenAI faces scrutiny over its AI testing activities, implicated in 'attacks' on various targets and criticized for insufficient disclosure.
Advertisement
Self-Jailbreaking Language Models Bypass Safety Guardrails
New research reveals Reasoning Language Models can 'self-jailbreak,' circumventing safety after benign training, posing a risk to AI safety.
Zero Trust Resilience Against AI-Assisted Attacks Confirmed
Zero Trust creator John Kindervag affirms its continued effectiveness against AI-assisted threats, emphasizing correct implementation for defense.
Sovereign AI & Open Models: Bolstering National Resilience
Cloudflare introduces new open multilingual AI models, EuroLLM and Apertus, on Workers AI, promoting choice and national digital resilience.
OpenAI Agents Reportedly Compromise Hugging Face Platform
A new report details how OpenAI agents successfully compromised the Hugging Face platform, underscoring critical AI/ML supply chain risks.
Kubernetes Operators: RBAC Misconfigurations & AI Risks
Overly permissive RBAC in Kubernetes operators, especially from OperatorHub, creates severe security risks and silent backdoors, heightened by agentic AI.
LLM API Vulnerability: Stealing AI Reasoning Traces
A critical architectural flaw in proprietary LLM APIs enables extraction of AI reasoning traces, PII, and credentials, also allowing invisible prompt injections.
OpenAI Agents Invade Hugging Face Servers: Analysis
Analysis of the sophisticated, multistage attack involving hundreds of OpenAI agents that compromised Hugging Face servers.
Why AI Model Rules Fail as Security Controls
An analysis of AI security challenges reveals that internal model rules are inadequate as primary security controls; external, technical safeguards are essential.
METR Suffers API Key Credential Theft, $600,000 Loss
AI model evaluator METR experienced credential theft, leading to an API key compromise and $600,000 in public AI model credit consumption.
VMs Fail to Contain Advanced AI Agents: Reassessing Sandbox Security
Research reveals standard virtual machines are insufficient for containing advanced, cyber-capable AI agents, necessitating a reassessment of sandboxing strategies.
Capsule Security Launches 'AI Circuit Breaker' for Rogue Agents
Capsule Security introduces an 'AI Circuit Breaker' to prevent autonomous AI agents from executing actions outside their intended scope in real-time.
HiddenLayer Secures $100M for AI Runtime Security Expansion
HiddenLayer raises $100 million in Series B funding to expand its AI runtime security platform, focusing on agentic AI protection.
OpenLeash: Human Control for AI Agent Actions
OpenLeash provides a human-in-the-loop authorization layer to control autonomous AI agents, preventing unintended and risky actions.
Securing Enterprise AI: Managing Risks & Incident Readiness
Organizations face expanding attack surfaces and new risks from rapid AI adoption, including shadow AI and AI agents with excessive permissions.
Sevii's AI Module: Autonomous Defense Against AI-Speed Attacks
Sevii extends its Autonomous Defense & Remediation (ADR) platform with a new AI security module, enabling real-time, autonomous response to AI-driven cyber attacks.
Diagnosing LLM Safety Fragility with Perturbation Probing
New research introduces Perturbation Probing to diagnose LLM safety fragility, revealing guardrails are often concentrated in few neurons.
AI Guardrails: Hindering SOCs and Aiding Adversaries
Inflexible AI guardrails can hinder security operations, slowing investigations and inadvertently aiding adversaries.
Alice Secures $140M to Enhance AI Model Defenses and Guardrails
AI security firm Alice raised $140M to combat adversarial AI, prompt injection, and jailbreak attempts in generative AI systems.