Pillar Security
Early-stage platform securing AI applications and infrastructure across the full AI lifecycle, from development through production.
Visit Website ↗Overview
Pillar Security builds guardrail and lifecycle-security tooling for AI applications, aiming to cover risk from initial AI development through production runtime. It positions itself as an end-to-end AI security layer rather than focusing on a single stage like red-teaming or posture management alone.
Founded in 2023 by Dor Sarig and Ziv Karliner, Pillar operates with a dual presence in Tel Aviv and Miami. It raised a $9M seed round in 2025 led by Shield Capital, with participation from Golden Ventures, Ground Up Ventures and strategic angels — a small round relative to several peers, reflecting its earlier stage.
Innovation Matrix Assessment
Still at seed stage roughly two years after founding, with development activity that has not yet been independently verified at scale.
Lifecycle-security positioning (development through production) is ambitious but the company is too early-stage for verified operational deployment evidence.
A $9M seed round is real but modest relative to peers in this category who have raised $18-180M; reflects genuinely early-stage momentum.
Full-lifecycle AI security framing is a reasonable extension of existing DevSecOps 'shift-left' thinking applied to AI, rather than a wholly new paradigm.
No independent efficacy evidence, named customer validation, or third-party benchmark was found for a company this early-stage.
Full-lifecycle AI security coverage addresses a real and growing need, though the category is crowded with better-funded competitors covering similar ground.
Why CISOs Should Care
Aims to give security teams a single control layer spanning AI development and production, potentially reducing tool sprawl for teams currently stitching together point solutions.
What Makes It Different
Positions itself around full AI-lifecycle coverage from the outset rather than starting narrow (e.g., red-teaming only or posture management only) and expanding later.
The Matrix Verdict
48/100 — EMERGING / UNRANKED
Emerging/Unranked: a real, seed-funded company with reasonable positioning, but too early-stage for meaningful independent evidence of efficacy or market traction relative to better-funded peers.
Editorial Note: Claims vs. Verified Findings
Funding details are corroborated across Calcalist, SiliconANGLE and Refresh Miami. No customer names, efficacy data, or headcount figures were independently verifiable at the time of this research; treat all platform-capability claims as UNVERIFIED vendor descriptions.
Sources
- SiliconANGLE — https://siliconangle.com/2025/04/16/pillar-security-raises-9m-address-growing-ai-specific-security-risks/
- Calcalist — https://www.calcalistech.com/ctechnews/article/hklv8g6cje
- Refresh Miami — https://refreshmiami.com/news/pillar-security-raises-9m-to-create-ai-security-guardrails-for-enterprises/
Alternatives to Pillar Security
Palo Alto Networks (Prisma AIRS)
Major cybersecurity incumbent's unified AI security platform (Prisma AIRS) for discovering, assessing and protecting AI models, agents and…
Zenity
Governance and security platform for AI agents and low-code/no-code development, securing agent identity, permissions and behavior across the…
Cisco AI Defense (Robust Intelligence)
AI security product line built on Robust Intelligence's algorithmic red-teaming and AI Firewall technology, now integrated into Cisco's…
Lakera
LLM security platform providing prompt-injection detection and real-time guardrails for generative AI applications, acquired by Check Point.
HiddenLayer
AI security platform unifying ML supply-chain scanning, runtime defense, posture management and automated red-teaming for enterprise AI systems.
Protect AI (now part of Palo Alto Networks)
AI/ML security company (model scanning, MLSecOps, AI-SPM) acquired by Palo Alto Networks for $634.5M and folded into Prisma…