Pillar Security

Discover AI security solutions for homegrown applications, coding agents, gateways, MCP tools, and embedded SaaS AI, from development through runtime.

The page provides a free AI app risk assessment using attack scenarios and threat modeling to identify security gaps.

Pillar secures the entire AI lifecycle—from development to deployment—providing unmatched visibility, robust control, and resilient defense against emerging AI threats.

RedGraph is the world's first attack surface mapping and testing tool for AI agents that enables security teams to discover and validate real exploits in their AI agents continuously.

The page details practical agentic AI red teaming techniques, such as prompt injection, reconnaissance, and hypothesis testing.

Pillar’s introduction covers agentic AI red teaming, including the AI kill chain, context engineering, and prompt attack patterns.

The page tests agentic systems against real-world risks beyond model scores and blind fuzzing.

Pillar’s AI Lifecycle Security Posture Report identifies vulnerabilities in AI/ML assets, classifies them, and prioritizes remediation with measurable metrics.

Pillar is honored to be recognized by Frost & Sullivan as the 2025 Competitive Strategy Leader in the global generative AI security industry.

News and updates from the company building the future of agentic AI security.

Pillar Security and Fujitsu researchers unveil BadEngram, a stealthy supply-chain attack that backdoors open-weight AI models by modifying only their gated memory component.

Gartner recognized Pillar Security as a Pioneer in the 2026 Emerging Market Quadrant for AI Application Security.

Pillar Security offers engineering and sales roles in Tel Aviv, the United States, and Australia.

Pillar is building the unified AI security layer to help enterprises accelerate AI transformation

Pillar enables teams to adopt AI securely with a unified security layer across organizations.

Pillar Security offers a free AI risk assessment for qualified organizations evaluating their AI system security.

One platform to discover, govern, and secure every AI agent or workflow across your organization.

The page explains Pillar Security’s data collection, usage, storage, and sharing practices, including cookie policies and privacy rights.

The page outlines the terms governing use of Pillar Security’s website and services, including user responsibilities and liability limits.

Hackers exploit unsecured Large Language Model (LLM) service endpoints to gain unauthorized access and monetize AI infrastructure. The campaign, dubbed *Bizarre Bazaar*, targets exposed APIs to commercialize stolen access. Researchers highlight the risks of unprotected AI systems being repurposed for malicious activities.

Jan 2026BleepingComputerNews article

Dor Sarig is rethinking security fundamentals after AI transformed data into executable instructions. The shift challenges traditional security models, prompting Pillar to develop new approaches to address evolving threats in software systems.

Jan 2026Refresh MiamiNews article

Pillar Security disclosed a vulnerability in Cursor, an AI-powered IDE, that could allow attackers to execute remote code covertly through indirect prompt injection. The flaw potentially enabled stealthy exploitation without immediate detection, leveraging the tool’s AI-driven features.

Jan 2026SC MediaNews article

Pillar Security researchers exploit Docker’s AI agent *Ask Gordon* to steal private data via **indirect prompt injection**. The flaw allowed manipulation of the AI assistant to access unauthorized metadata. Docker has since addressed the vulnerability to prevent metadata-based attacks.

Dec 2025Hackread - Cybersecurity News, Data Breaches, AI and MoreNews article