What's Your AI Liability Right Now?

Our AI agent analyzes your inputs through a multi-step pipeline: it filters relevant risks from our taxonomy based on your system profile, ranks them by potential exposure, estimates financial impact using Fermi estimation (frequency × loss per event), validates the reasoning for consistency, and generates actionable narratives.

We use Fermi estimation grounded in epistemological rigor. For each risk, we estimate how often incidents might occur (frequency) and the financial impact when they do (fines, legal costs, remediation, indirect losses). We adjust base rates from industry precedent for your specific context—jurisdiction, company size, automation level, and data sensitivity.

Every input matters. Your deployment type determines which risks apply. Jurisdiction affects regulatory exposure (EU AI Act fines can reach 6% of revenue). Company size scales absolute costs. Automation level affects risk magnitude. Data sensitivity (health, biometric, children's data) triggers specific regulatory requirements. The optional description and URL help us understand your specific context.

About 3 minutes to complete the form, then 5-7 minutes for our AI agent to analyze. You'll see real-time progress as each step completes—the agent filters relevant risks, estimates financial impact for each, validates its reasoning, and generates your personalized report.

You receive an expected annual loss estimate with 90% confidence intervals, broken down across five risk categories: Technical, Operational, Legal & Compliance, Ethical & Reputational, and Governance. Each category includes key risk drivers, assumptions, and suggestions for what would reduce your exposure.

Risk estimation is inherently uncertain. A single number creates false precision. Our 10th-90th percentile range tells you: "In most scenarios, your exposure falls within this band." The expected value is our best estimate, but the range shows realistic uncertainty.

Technical: Model failures, data quality issues, system vulnerabilities. Operational: Process breakdowns, human errors, integration failures. Legal & Compliance: Regulatory fines, contractual breaches, litigation. Ethical & Reputational: Bias incidents, privacy concerns, brand damage. Governance: Oversight gaps, accountability failures, documentation deficiencies.

No. This is an informational estimate based on general risk models, not legal advice, audit opinion, or regulatory guidance. For binding assessments, consult qualified legal and compliance professionals.

We estimate financial loss over a one-year horizon. We don't cover: existential AI risks, geopolitical impacts, multi-year compounding effects, or harms that can't be quantified financially (like privacy violations beyond regulatory fines). We also can't assess your actual governance maturity—we assume a moderate baseline.

The estimates are directionally useful, not precisely predictive. We use industry base rates and adjust for your context, but we don't have access to your internal data, incident history, or actual governance practices. Think of it as a structured expert estimate, not a guarantee.

Key assumptions include: (1) your organization has moderate AI governance maturity (not excellent, not poor), (2) base rates from industry precedent apply to your context, (3) regulatory enforcement patterns will continue similarly. All assumptions are surfaced in your results.

Yes. The assessment is free with no login required. We offer this to help organizations understand their AI risk exposure and to introduce Modulos' AI governance platform.

Yes. Your results have a unique URL you can bookmark or share. You can also download a PDF report.

Use them to prioritize which risks need attention, inform budget conversations about AI governance, and identify where controls or oversight might reduce exposure. For high-risk systems, consider a formal assessment with governance experts.

Yes. When you provide a URL, we extract context about your company, product, scale, and any compliance claims. This helps us tailor estimates beyond what the form fields capture. Without it, we rely solely on your form inputs.