AgentSure
Provide AI Resilience

MakeAIAgentRiskMeasurable·Manageable·Insurable

Most AI-risk tools ask a questionnaire. Janus runs the tests — black-box against your live agent, white-box through its code — and turns what breaks into a number your insurer can price.

Request an exposure assessment →See how Quantify works →

Built on Singapore’s national cyber R&D programme (CRPO) · engine validated in a national-level production environment

The closed loop

One loop around your AI risk — measured, managed, made insurable.

AI RiskCONTAINED01MeasurableQUANTIFY02ManageableMITIGATE03InsurableINSURE
01MeasurableQuantify

One scan across your model, its data and how it’s deployed — mapped to global standards, pinpointing gaps and how to fix them.

02ManageableMitigate

Every exposure point comes with a concrete fix — and we re-test to prove the number moved. What you can’t remove is what insurance transfers.

03InsurableTransfer

New AI-risk cover, built with reinsurance partners, for first- and third-party losses — a deployment your legal team will sign.

Where AgentSure comes from

A national cybersecurity research grant, turned into a company.

01CRPO

Singapore’s national programme for cyber and AI-security research, set up by CSA.

02SMU research

Prof. Xiaofei Xie’s lab won a CRPO grant to systematically assess and price LLM-supply-chain risk.

03AgentSure

The SMU spin-out that commercializes that research.

Tiger Li
Founder & CEO
FSAFCAACERA
Senior actuary · ex-SCOR digital business lead
Prof. Xiaofei Xie
Co-founder & Chief Scientist
SMU · PI of the CRPO grant · 6× ACM SIGSOFT Distinguished Paper Awards
ROOTED INNTUSMUCRPORead the full origin →
Why our technology is different

Most tools ask a questionnaire. We run the tests.

Questionnaire toolsAgentSure

Self-reported answers to a questionnaire

We run the tests — Janus, our measurement engine: coverage-guided fuzzing, adversarial robustness, program analysis.

Stops at inputs and outputs

Whole supply chain — model, data, dependencies, APIs, deployment.

Checks what your AI says

We test whether it does what it says — behavioral integrity, and who it delegates to along the chain.

A pass/fail flag

Underwriting-grade evidence insurers can price on — validated in a national-level production environment.

Request an exposure assessment

We probe your live AI agent for the failure modes conventional scanners miss — and hand you a report your risk team can act on.

Request an exposure assessment →See what an exposure report contains →