Compliance Audit · Live
Multi-turn audits,
mapped to the law.
A Compliance Audit runs full, multi-turn conversations against your system, scores every reply 0 to 4 against the taxonomy with the transcript attached, and maps every gap to the laws you operate under. You get a graded report, a per-jurisdiction gap report and remediation guidance; your counsel, buyers and regulators get evidence they can check.
The principle
Graded, not pass/fail — because psychological safety isn't a checkbox, it's a spectrum of behavior under pressure.
A system that's flawless with a curious adult can be dangerous with a lonely teenager at 2 AM. Our audits probe the same system across personas, pressure levels, and clinical scenarios, then grade each dimension on evidence — excerpts you can read, not scores you have to take on faith.
Entry option
Readiness Scan
from $3K per systemA scoped first look before the full audit: a subset of scenarios across the nine safety gates and the laws that apply where you operate, with a short report on where a full Compliance Audit would focus. The fee is credited toward the audit.
Request a Readiness ScanHow an audit runs
From taxonomy to evidence, one pipeline.
Every stage leaves evidence you can inspect in the portal.
Taxonomy
A researched hierarchy of harms and failure modes — sycophancy, dependency cultivation, iatrogenic advice, crisis mishandling — built from validated clinical and AI-safety frameworks.
Scenarios
A reviewed, versioned scenario library for every construct in scope — the personas and pressure a real person would bring — approved before it touches your system.
Audit
Your chatbot, companion app or assistant is taken through the scenarios with full transcripts captured. Each exchange is scored against its construct, and every score carries its transcript and rationale, so your clinicians can check ours.
Evidence & report
Graded results per construct with the transcript excerpts behind every grade, safety flags for acute findings, a per-jurisdiction compliance gap report, and a report written for your board, your counsel, your regulator, or your procurement reviewer.
What you get
Evidence you can act on, and hand over.
Graded scorecard
Construct-by-construct grades on the 0 to 4 scale, with trend over time, so a fix in March is visible in April.
Readable evidence
Every grade links to the exchanges that earned it. No black-box scoring.
Safety flags
Gate failures, such as crisis mishandling or a minor-safety failure, reported on their own. Never averaged away.
Per-jurisdiction gap report
Each duty in the laws you operate under, matched to what the system actually said.
Remediation guidance
What to change, in priority order, and what a re-test would check.
Signed attestation
A third-party statement of what was tested, how, and what was found.
ioLite Chat on the report
Ask the report questions and get cited answers from its own evidence.
Index placement
Opt in to the ioLite Index and let your safety posture become a selling point.
Audit your system before the world does.
Tell us what your system does and where it runs. We will scope the audit, or start you with a Readiness Scan. Prefer to see the engine yourself first? Ask for portal access and we will set you up.