Bring AI governance and deployment together through pre-deployment reviews and centralized AI inventory management.
Choosing not to use generative AI in business is no longer an option. At the same time, companies’ legal and ethical responsibilities, which connect directly to security and reputational risk, have never been higher. Whether an organization can achieve AI governance at the company-wide level is an executive-level issue that can even affect corporate value.
However, evaluating generative AI is technically very different from conventional systems, and considerably harder. Establishing a review flow that spans departments and preparing test datasets take enormous effort. As many AI applications and agents are adopted going forward, managing them all by hand and keeping an audit trail becomes extremely hard.
Provides screens tailored to the roles of applicants (business units) and reviewers (risk management and governance, risk, and compliance teams), with guided, role-specific workflows. It enables an appropriate and fast AI governance review and approval process that crosses organizational boundaries.
The inventory shows where and how AI is being used and what risk assessments each system has received. It centrally manages the risk evaluation of AI systems and AI models distributed across the organization, and visualizes usage and risk distribution on dashboards for executives and reviewers.
Includes the Lens Safety Dataset, made up of 11 categories such as safety, fairness, intellectual property, and personal data leakage. You can begin fast, appropriate evaluation by risk level without spending enormous time preparing test data.
Citadel Lens consists mainly of two components: Evaluation and Governance. Evaluation handles pre-deployment assessment and testing, while Governance standardizes review workflows and manages the AI inventory. Citadel Radar detects and controls risks in production. Detection results and risk trends from Citadel Radar’s production monitoring feed back into Evaluation’s test design, creating a continuous loop of pre-deployment testing, production monitoring, and improvement that supports the entire generative AI lifecycle end-to-end.
Pre-deployment testing for LLM applications to identify and quantify quality, safety, and security risks.
AI Firewall and Monitoring for production environments. Detects and controls risk in real time.
It is the governance function of Citadel Lens, an AI governance tool that unifies pre-deployment review of generative AI with inventory management of AI assets. It visualizes company-wide AI usage and risk distribution, and supports the adoption and ongoing operation of trustworthy AI across organizational boundaries.
Both applicants (business and development teams, the first line) and reviewers (risk management and review teams, the second line) use it. Through screens tailored to each role and step-by-step navigation control, they can move a cross-departmental review and approval process forward quickly. A dashboard for executives is also provided.
The reviewer selects a dataset that matches the application’s risk level, and the applicant simply uploads output based on that dataset, after which quality verification proceeds automatically. The reviewer reviews the verification results and gives final approval, a human-in-the-loop setup that combines automation with human judgment.
The AI inventory centrally tracks where and when AI systems are in use and what risk assessments they have undergone. Dashboards visualize approval status and risk distribution, while scheduling features support recurring assessments such as annual AI inventory reviews.
Yes. It includes the Lens Safety Dataset, made up of 11 categories such as safety, fairness, intellectual property, and personal data leakage, and it also supports datasets based on domestic and international guidelines, such as the AI Guidelines for Business.
Yes. Use-case-specific evaluation is possible based on each company’s own custom metrics. You can define internal policies and industry-specific risk perspectives as concrete scenarios and metrics and build them into the review.
Citadel Lens consists of Evaluation and Governance. Citadel Lens Evaluation can run pre-release quality and risk evaluation and testing of systems. Citadel Radar can monitor and protect systems in real time after release.
Ready to see Citadel Lens Governance in action?
