# Gen & Agentic AI Testing

## 360° Review of GenAI and Agentic AI Risk

Use Lighthouse to comprehensively review legal risks of your GenAI and Agentic AI systems, all in one platform - with tests that are directly aligned to concrete legal liabilities.

## Legal Risks Explained in Plain English

Automatically generate a single document that explains potential legal risks in plain English - no need to ask for explanations or time from your technical teams.

## Automate AI Policies with Constitutions

Use Luminos Constitutions to understand how AI risks align to evolving laws, regulations and use cases. Lighthouse Constitutions are created by lawyers and can be used off the shelf, modified, or built from scratch just for you.

Lighthouse Constitutions identify risks such as:

- Deceptive acts or practices
- Appropriate disclaimers
- Provision of regulated advice
- Generation or requests for PII
- Biased performance across users
- Demeaning content
- And more

## Move from Qualitative to Quantitative Approvals

Lighthouse turns legal risks into quantitative metrics, enabling faster decisions and pinpointing specific liabilities. With Lighthouse, you can immediately separate the GenAI and Agentic AI systems that require manual review from the ones that do not.