Understanding AI Agents - The Next Wave of Automated Internet Activity Artificial intelligence is reshaping not only what software can do, but how it operates across the internet. While in the recent past, AI’s most visible impact on web activity came in the form of AI crawlers that collected and indexed content at scale for model training, a more complex shift is underway. Dhanesh Ramachandran |March 17, 2026
How Effective Is Radware’s AI Agent Protection Against Indirect Prompt Injection? As autonomous AI agents rapidly enter business workflows, security teams face a new and often overlooked threat: indirect prompt injection (IPI). Dror Zelber |February 18, 2026
When Good AI Agents Go Bad: A Cautionary Tale for Modern Organizations In 2025, retail chain BrightMart was riding a wave of optimism. Like many organizations across industries—law firms, insurance companies, logistics providers and customer-facing service companies—it had begun deploying AI agents to accelerate employee productivity, automate repetitive tasks, streamline customer interactions and push overall efficiency to new levels. Dror Zelber |February 03, 2026
When Help Turns Harmful: How Attacking a Healthcare LLM Prompt Can Put Patients at Risk Healthcare institutions around the world are adopting AI-driven virtual assistants to improve patient services. Instead of waiting on hold, patients can ask a Large Language Model (LLM) for help with booking appointments, checking lab results, understanding treatment options, managing chronic conditions, or even getting reminders about medication or follow-ups. Dror Zelber |January 14, 2026
ZombieAgent: New ChatGPT Vulnerabilities Let Data Theft Continue (and Spread) We’ve discovered several new vulnerabilities that allow an attacker to exploit ChatGPT to exfiltrate sensitive or personal information. Zvika Babo |January 08, 2026
From Cutting-Edge to Critical Risk: Unpacking the Cybersecurity Dangers of LLM Integration - Part 2: Defending at the Inline Edge In Part 1, we explored how integrating large language models (LLMs) into business applications creates new and often misunderstood security exposures - from prompt injection and data leakage to brand impersonation and compliance risks. Rotem Elharar |December 16, 2025
Top 5 LLM Security Risks Every Business Must Address The adoption of large language models (LLMs) is revolutionizing how businesses operate, from automating customer support to accelerating content creation and data analysis. These models offer unprecedented efficiency and personalization, giving companies a competitive edge and enhancing customer experiences. Dror Zelber |November 18, 2025
ShadowLeak: A Zero-Click, Service-Side Attack Exfiltrating Sensitive Data Using ChatGPT’s Deep Research Agent We found a zero-click flaw in ChatGPT’s Deep Research agent when connected to Gmail and browsing: A single crafted email quietly makes the agent leak sensitive inbox data to an attacker with no user action or visible UI. Co-Lead Researchers: Zvika Babo, Gabi Nakibly; Contributor: Maor Uziel |September 18, 2025
From Cutting-Edge to Critical Risk: Unpacking the Cybersecurity Dangers of LLM Integration – Part 1 Large Language Models (LLMs) are reshaping industries, unlocking unparalleled innovation and efficiency. But with this progress comes a serious concern - new cybersecurity risks that organizations must confront. Rotem Elharar |May 21, 2025