What’s new: Openlayer named in the 2026 Gartner Market Guide® for AI Evaluation and Observability Platforms. Learn More
All customersCybersecuritySince 2023

How a cybersecurity innovator kept phishing attack simulations 95%+ compliant using Openlayer

Jericho SecurityPublished March 27, 20255 min read
#01

95%+pass rate

across phishing email outputs

#02

6+months

of engineering effort saved

#03

50%faster

prompt iteration cycles

Situation + Impact

Jericho Security operates in the cybersecurity space, providing enterprise-grade phishing simulations and employee training. By leveraging GenAI, Jericho creates realistic email, text, and voice scenarios that strengthen employee security practices.

As adoption accelerated, Jericho Security faced a common challenge with generative AI at scale: ensuring that outputs remained both realistic and compliant with downstream validation requirements. For example, phishing emails had to follow precise formatting rules to enable proper link validation. Failures introduced operational inefficiencies, and without visibility into LLM performance, the team could not confirm whether prompt updates improved outcomes or introduced regressions.

Building internal monitoring would have taken months of engineering resources. Instead, Jericho deployed Openlayer to provide continuous oversight and guardrails, ensuring reliable outputs, faster iteration, and scalable governance without compromising velocity.

Without Openlayer, we’d be blind to how our LLM outputs behave at scale. It’s saved us months of engineering time and given us a repeatable way to keep phishing simulations reliable and realistic
Daniel Chyan — CTO, Jericho Security

Results

With Openlayer Monitoring Mode, Jericho Security aggregates and analyzes large volumes of outputs, automatically detecting formatting or accuracy issues that could undermine phishing simulations. This prevents regressions and enforces consistency across enterprise-scale deployments.

Development Mode enables Jericho’s engineers to experiment with new prompts and benchmark prompt sets to evaluate realism and believability. These capabilities reduced prompt iteration cycles by 50% and gave teams confidence that updates improved results.

As a result, Jericho consistently sustains >95% pass rates across email phishing outputs, with fewer than 1 in 20 requiring re-generation. By embedding Openlayer directly into its CI/CD pipeline and daily engineering workflows, Jericho has avoided over six months of engineering effort while gaining the guardrails and visibility required to deliver secure, realistic, and compliant phishing simulations, outpacing legacy competitors.

Write the next story.

See your first AI system tested, monitored, and audit-ready in one demo.

2026 Openlayer. All rights reserved.