Blog Asaf Saar

Asaf Saar EVP Product

Asaf Saar is EVP and Chief Product Officer at Mend.io, where he leads product strategy for the company’s application security and software composition analysis platform, including its work securing AI-generated code and AI components. He joined Mend.io after more than five years as VP of Product Management at Tricentis. Asaf has spent his career building and leading product organizations at scale, with a focus on developer tooling, quality, and turning technical depth into commercial momentum.

OWASP LLM Top 10 2026: the model will be fooled, the question is what breaks - Featured image OWASP LLM TOP 10 2026

OWASP LLM Top 10 2026: the model will be fooled, the question is what breaks

Asaf Saar Aug 13, 2026

AI Models Risk

The 2026 OWASP LLM Top 10 shifts the job from prevention to containment.

OWASP LLM Top 10 2026: the model will be fooled, the question is what breaks - Featured image Move Faster Than AI Driven Risk 1000x650

Move faster than AI-driven risk: Inside Mend.io’s latest AI application security update

Asaf Saar Jul 28, 2026

AI Models Risk

AI agent discovery, runtime guardrails, agentic triage, and zero-day speed.

OWASP LLM Top 10 2026: the model will be fooled, the question is what breaks - Featured image Independence Is the Moat

Independence is the moat

Asaf Saar Jul 14, 2026

AI Models Risk

Why the independent security layer keeps winning as AI models get better.

OWASP LLM Top 10 2026: the model will be fooled, the question is what breaks - Featured image When the Guardrails Held and the Attack Still Worked 1000x650

When the guardrails held and the attack still worked

Asaf Saar Jun 30, 2026

AI Models Risk

The guardrails held. The AI-assisted attack still worked. Here is why.

OWASP LLM Top 10 2026: the model will be fooled, the question is what breaks - Featured image AI Changed What You Ship 1000x650

AI changed what you ship. It also changed what you have to secure.

Asaf Saar Jun 22, 2026

AI Models Risk

AI changed what you ship and what you have to secure.

OWASP LLM Top 10 2026: the model will be fooled, the question is what breaks - Featured image The Wrong Meter 1000x650 e1782464489919 1

Frontier Model Is the Wrong Meter for Continuous Security

Asaf Saar Jun 18, 2026

AI Models Risk

Why frontier model security is too costly to run as an always-on scanner.

OWASP LLM Top 10 2026: the model will be fooled, the question is what breaks - Featured image The Verification Layer Cannot Be the Model 1000x650 e1782465074756 1

Why AI Can’t Verify Its Own Code and What That Means for Enterprise AppSec

Asaf Saar Jun 16, 2026

AI Models Risk

AI-generated code security needs an independent verification layer.