validant.ai builds systems for trust and agency. For any AI system that decides about people, it runs one workflow: assess the system, communicate the results to everyone who relies on them, and validate them as proof anyone can check. It covers three bodies: the model, the organisation behind it and the person it decides about. Use the platform yourself, or let us run the workflow with you.

We sign the measurement: a seal attests the results our engine computed, by the stated method, on the data supplied. It is not a legal verdict and not a promise about any decision or its outcome. That conclusion is for your counsel, or for a partner auditor where a law asks for one.

The platform comes in self-serve plans. When we run the workflow with you, with advice on what to change and how to govern it, it is part of an Enterprise subscription, scoped per AI system and priced per contract against the audit work it replaces, never on whether a system passes. A free first reading gets you started; pilots are paid.

What does validant.ai do for AI fairness?

It runs end-to-end AI fairness and explainability audits: detecting bias and proxy variables, mitigating disparities across data, models, LLMs and agents, explaining decisions with SHAP, LIME and counterfactuals, and monitoring fairness in production with auditor-grade, regulator-ready reports.

Who is validant.ai for?

Anyone who wants AI to stay fair, understandable and under human control. That includes risk, compliance, legal and ML teams proving an AI system is fair, explainable, trustworthy and accountable under the EU AI Act, NIST AI RMF and ISO/IEC 42001, and just as much the everyday person who wants to safeguard their own data and keep control of their agency. We deliberately keep it plain enough for everyone to use, not only specialists.

What is Verifiable Trust Infrastructure (VTI), and how does validant.ai relate to it?

VTI is an open specification, currently a working draft, developed in the Decentralized Trust Graph Working Group run jointly by Trust over IP and the Decentralized Identity Foundation. It defines how people, communities, organisations and the AI agents acting for them vouch for each other, delegate authority and revoke it without a central gatekeeper. validant.ai's founder is a contributor at both foundations and active in that working group, and validant.ai's seals, decision receipts and agent mandates are designed to plug into VTI as it matures.

01INTRO

Trustworthy AI, within everyone's reach

Make AI fair, explainable and provable: make AI trustworthy.

We test it, explain it, and seal the result as proof anyone can check.

Contributing to

Ecosystem

02HOW IT WORKS

How it works

Fairness, assessed.Results, communicated.Proof, validated.

The same three steps for a hiring model, a credit score or an AI agent. Each one ends in evidence someone else can check, not a slide someone has to believe.

Assess
We test whether an AI system treats groups of people differently, across 47 fairness metrics and statistical tests, with the evidence to show how sure we are.
You getA plain-language verdict, with the numbers behind it.
Communicate
We show which factors drive each decision and by how much, and say it in words each reader needs: counsel and the board, a regulator or works council, and the person affected.
You getA report for every audience, and an explanation for the person.
Validate
We seal the measurement as a signed credential. Anyone can check it is authentic, unaltered and not revoked, with no login and without trusting us. It covers one system at one moment, not every decision forever. We sign the measurement; your counsel signs the conclusion.
You getA Trust Seal for the system, a Decision Receipt for the person.

How it works, in figures

47fairness metrics and statistical tests
3steps: assess, communicate, validate
1seal anyone can check, without an account

Illustrative data

A model scores 40 applicants from each of two groups with the same spread of qualifications. Move the shortlist line and watch who gets through.

Group A

19 of 40 shortlisted

Group B

7 of 40 shortlisted

Shortlist everyone scoring at least58

Selection-rate ratio

0.37

Group A 48% · Group B 18%. Group B is shortlisted at 0.37× the rate.

What drives the gap

Postcode. It stands in for group membership and costs group B points it did not earn.

A result that fails is not sealed.

The four-fifths rule is one screening test, not the whole fairness question. A real audit runs 29 modules with confidence intervals, not one ratio.

The same decision, followed across all three bodies, with nothing asserted that a stranger cannot check. It is the shortest honest answer to what we build, which is why we use it as our worked example rather than a diagram.

The model. A CV screening model is measured, and the seal's subject is that AI system at that moment. The engine signs the numbers; our server derives the verdict from them.
The organisation. The deployer is named on the credential as accountable owner, next to the issuing identity and legal entity. It cannot assert its own result.
The person. The applicant gets a Decision Receipt in her own wallet instead of a form rejection: why, the decisive factor, her route to human review, and a link back to the seal she can check herself.

How the seal works Check a seal yourself

03USE CASES

Use cases

One workflow, every decision about people.

Each use case adds its own tests, law and readers to the same three steps.

Most finishedBeta
Hiring and HR. Role by role.
Test every role, not the company average, and give each candidate a receipt and a route to a human.
Four-fifths rule · EU AI Act Annex III · NYC Local Law 144
Read the hiring guide
Planned
Credit and lending
Explain every score-led decision the way the law asks.
GDPR Art. 22 · SCHUFA ruling
Planned
Insurance
Show that pricing and underwriting treat groups fairly.
EU AI Act Annex III · pricing disparity
Planned
Health and public sector
Evidence at the level of each decision, in words people can read.
Clinical and public decisions
Planned
Generative AI and agents
Test what models say and what agents do on your behalf.
Prompt fairness · signed mandates

Your use case is not here? The workflow is the same.

04ASSURANCE

Three bodies · one assurance

Every AI decision touches three parties. We assure all three.

The model that decides, the organisation that answers for it, and the person it decides about. Trust holds only when all three can be checked, never one alone.

Why trust is an orbit, not a pillar

Editorial illustration of a long colonnaded courthouse hallway in perspective, with a single figure walking toward the vanishing point

Metaphor

The hall of due process

Fairness in ML is not a feeling, it is a corridor of evidence. Each column is a fairness metric the model must pass through before it reaches a decision. The figure walking toward the vanishing point is the model under audit.

the model

Live

fair.AI fairness & explainability

What your model actually does, with the receipts to prove it.

Detect, mitigate, explain and monitor bias across the whole pipeline, from training data to LLMs. SHAP, LIME and counterfactuals turn “the model is unfair” into “this feature drives it, and by this much”, with the statistics to back it and EU AI Act mapping built in.

29 assessment modules across the full lifecycle
43 discrimination patterns · US / EU / CH jurisdictions
Explainable AI built in · SHAP, LIME and counterfactuals
Bootstrap, Bayesian and permutation testing

See in the platform

Metaphor

The Swiss Federal Palace

Civic institutions are where digital trust either holds or breaks. We chose Bern because Switzerland is where Validant is built and where digital sovereignty is taken as a serious craft, not a slogan.

the organisation

Service today

accountable.Digital trust

Proof that your organisation keeps its promises.

Built on the iceberg.digital framework, we audit whether your digital systems behave as promised: in governance, in engineering, and for the people who use them. Most of that work happens below the waterline, which is where trust is actually won or lost.

Ten trust constructs · 128 measurable cues
Governance, engineering, agency and institutional layers
Explicit and implicit signals, continuously assured

We support digital-trust audits today, as a service. The platform modules are in progress.

Editorial illustration of a hand holding up a teal ID card against a city skyline, evoking self-sovereign identity

Metaphor

The credential in your hand

Identity should fit between your fingers, not in someone else's database. The hand is yours. The card is the proof. The city behind it is everywhere it can be presented and verified, with no platform in between.

the person

Planned

yours.Identity & agency

A record of every AI decision, held by the person it is about.

A Decision Receipt gives the person affected their own copy: why the decision fell as it did, and how to ask for a human review. Next come portable credentials and signed agent mandates, so an agent acting for you can prove what it may do, and you can take that back.

Decision Receipt in the person's own wallet
Trust Experience Library of first-person patterns
Credentials and signed agent mandates on the roadmap

05FILMS

Films

Where an algorithm already decides.

Four short films, four places where AI already makes the call: what a child is shown, who gets the job, who gets the loan, who gets care. Pick a use case.

Choose a use case

Hiring · Full film · 2:41

Trust me, I'm an algorithm. A self-driving car that has read the entire internet, mostly the comment sections, screens Maria's job application. Postcode. Rejected. You can outsource the decision. Not the responsibility.

Share this film

06STANDARDS

Open standards · VTI

An open trust layer, for people and the AI that acts for them.

Whether a person proves who they are, an organisation answers for its AI, or an agent acts on someone's behalf, the proof is only worth something if anyone can check it. That takes shared infrastructure, not one more platform.The Verifiable Trust Infrastructure (VTI) is that layer, being specified in the open by Trust over IP and DIF. Our seals, receipts and mandates are designed to plug into it as it matures.

Read the VTI draft specificationWorking draft 0.2 · not yet ratified

01Where trust is needed
People, organisations and their AI
A person proving who they are, a company answering for its model, an agent acting for someone.
02What we build
Trust Seal · Decision Receipt · Agent Mandate
The proofs we build: a result anyone can verify, a receipt the person keeps, authority an agent can show.
03Where our founder contributes
Verifiable Trust Infrastructure
How people, communities, organisations and agents vouch for each other, delegate and revoke, with no gatekeeper.
04Open foundations
W3C DIDs and Verifiable Credentials
The open identifiers and credential formats everything above is built on.

Checkable by anyone

A proof is only useful if a stranger can verify it. Shared infrastructure means no single company decides who is trusted, us included.

People stay in control

A person holds their own credentials, vouches for others, and can hand authority to an agent and take it back, without a platform in between.

Built in the open

VTI is specified in the joint Trust over IP and DIF Decentralized Trust Graph Working Group, where our founder is a contributor.

07PLATFORM

The Validant platform

One platform, three bodies, every claim auditable.

Fairness, digital trust and identity run as modules on one platform. Pick what you need, share the evidence, and keep every claim open to inspection.

What you get.

What we will not negotiate

Five commitments, built into the runtime.

Standards before features
W3C Verifiable Credentials, DIDs and the open work at Trust over IP and DIF come first, with MLflow and pytest hooks for the tools you already run. We extend ecosystems, we do not lock them in.
Rigour a peer reviewer accepts
Bootstrap and Bayesian intervals, permutation testing, effect sizes, multiple-testing correction. Every claim survives a peer reviewer reading the table.
Compliance is the default
EU AI Act risk classification, GDPR-aligned privacy tiers and 43 historical discrimination patterns across US, EU and Swiss jurisdictions, wired into every assessment.
Auditable in production
Real-time drift detection, a fairness gate before release and four-eyes validation. What an agent does on your behalf is logged, signed and revocable.
Privacy and sovereignty by default
Three privacy tiers per dataset: k-anonymity, differential privacy or exact values. Credentials and mandates are held by the person, never by us. No silent telemetry, no shadow profile.

Open the platform

08WORK WITH US

Work with us

Three ways in, one workflow.

Use the platform yourself, or let us run the workflow with you and advise you on the way.

The platform
Do it yourself.
- Assessments, explanations and sealed reports
- Your analysis data encrypted in your browser
- Invite-only early access
Trust Evidence · Enterprise
We run it with you.
- We assess, communicate and validate your AI system
- Advice on what to change and how to govern it
- A re-check every cycle, a seal anyone can check
We sign the measurement; your counsel signs the conclusion.
With partners
Their judgement, our evidence.
- Law firms and consultancies sign the conclusion
- Testing bodies keep their mark
- One evidence file every reader can check
For partners

A free first reading gets you started. Pilots are paid. Trust Evidence is part of the Enterprise subscription.

09PARTNERS

Partners

You sign the conclusion. We supply the signed evidence.

For law firms, consultancies and testing bodies whose clients use AI that decides about people.

Law firms: Advise and sign the legal conclusion, on evidence your client can check.
Consultancies: Run the governance work; we supply the measurement under it.
Testing bodies: Keep your mark. Our signed evidence is the measurement behind it.

10LIBRARY

vfairness · open libraryBeta

The fairness library that powers the platform.

vfairness is the engineering core inside validant.ai: a Python library for measuring, mitigating, explaining and monitoring fairness across the whole pipeline, LLMs included. Tap any number below for a plain explanation.

Read the library reference Explore the 29 modules

One book, a thumb tab for each of its twelve parts

The reference work behind the libraryPublic · CC BY-NC 4.0

The AI Trust & Fairness Compendium

Twelve standalone parts with 115 fixed chapter numbers and three appendix volumes, written from identified sources, reviewed and fact-checked. One notation and one vocabulary for everything the library measures and the platform assesses.

12
parts: 115
chapters: 933
defined terms

Read the compendium

11SIGNAL

News & updates

Signal.

What's new at validant.ai — product films, releases and notes from the work. The latest, in order.

View all

Fine grey-blue technical drawing on warm paper of a large wheel of Gruyère cheese on a spruce board, drawn in isometric. On the rind sits a small teal oval casein mark of abstract dots and dashes, shown enlarged in one section call-out circle, and a small human figure beside the wheel holds a magnifying lens up to it. From the mark a single fine dashed line sweeps across open paper and gentle pasture contours, past one small cow, to a small village cheese dairy whose cut-away wall shows a copper vat. A soft coral wash sits behind the wheel.

Featured · Events

29 October 2026Daniel Glinz15 min read

You Cannot Audit What You Cannot Trace

A fairness verdict rests on data twice over: the data a model learned from and the data it was tested on. Notes from the Swiss Data Space Forum 2026 and the Swiss Data Alliance white paper on AI-ready data, on why data spaces and fully open models decide what an assessment can prove.

Read the report

Fine grey-blue technical drawing on warm paper of an open-wheel racing car braking into a curved corner, drawn large in isometric. The near front wheel is cut away to show a ventilated brake disc gripped by a teal brake caliper, repeated in a section call-out circle as an exploded detail. Dimension arcs trace the corner, blank boards line the track, and fine dashed telemetry lines run from the car over the trackside barrier to a lectern outside the track, where one small human figure reads an open ruled ledger. A soft coral wash sits behind the front of the car.

ResearchOpen to read

5 October 2026

Controlling Agents Triples Their Value. Proving It Is What Makes It Count.

BCG surveyed 1,330 executives and found that companies which control their AI agents properly get about three times as much value from them. Only 5% have those controls in place, and every one of the six is checked by the company itself. A reading of the Applied AI Index 2026, the gap it leaves open, and why proof is what lets the tripled value count.

Read

Fine grey-blue technical drawing on warm paper of a large water tank whose walls are built from corrugated motorway crash-barrier rails bolted to steel posts on a concrete plinth, with fish swimming calmly inside. One rail is still missing from the near wall, and a crane lowers the teal replacement section toward the gap while a worker on a scaffold guides it in and a second worker signals from the ground. A single fish has already passed out through the unfinished gap along a dashed arc, with a soft coral wash behind the tank.

EventsOpen to read

28 September 2026

The Preemptive Assurance Paradox: Who Guards the AI Guardrails?

In one September week a lab asked to slow down, a president declared himself the only guardrail AI needs, and The Economist asked whether the arms race can be stopped. Notes from Trust Valley Days 2026 at EPFL on why the answer is not a pause or a sprint, but liability, outcomes and verification.

Read

Fine grey-blue technical drawing on warm paper of an astronomer at a large telescope inside a single bounded circular aperture ring: a few stars rendered sharp and resolved within the ring, many fainter stars beyond it washed in pale coral, the aperture ring picked out in teal, and a small human figure at the eyepiece for scale.

ResearchOpen to read

10 August 2026

Precision Is Not Proof: The Trap in Every AI Fairness Verdict

A model that predicts perfectly stops insuring anyone; a verdict published without its detection limit invites you to assume the limit is zero. Two failures, one omission, and two borrowed words for keeping them apart.

Read

Swiss-style line drawing of Daniel Glinz smiling, in a light beige blazer and open-collar shirt with a conference lanyard, holding a certificate that reads Best Paper Award, The Architecture of Digital Trust, on a clean white ground.

ResearchOpen to read

24 July 2026

The Trust Problem Nobody Wants to Name

Everybody is spending on AI; almost nobody can say what they got back. A new paper, awarded the Best Paper Award at the IEEE Swiss Conference on Data Science and AI, names that gap and shows it is a trust problem wearing a technical costume: seven mechanisms, a four-layer architecture, an iceberg beneath it, and five design principles for closing it.

Read

Fine grey-blue technical drawing on warm paper of a single rack-mounted frontier-AI model unit gone dark, beside a large knife-blade breaker switch thrown to off and actuated by a sealed letter, with dashed data lines to small client terminals and a faint world map all cut by break-marks, one teal switch lever, a soft coral wash and a small human figure for scale.

ResearchOpen to read

14 June 2026

When a Model Becomes a Munition

A frontier AI model was switched off worldwide by a single government letter. Reading the June 2026 shutdown through the three-body picture of digital trust: how the kill switch stopped being a metaphor, why the most governable lab was governed least carefully of all, and what single-vendor, single-jurisdiction dependence now costs a board.

Read

Fine grey-blue technical drawing of an orrery: a heavy central sphere in a teal gimbal mount, encircled by calibrated elliptical orbit rings with two smaller spheres captured close, a key hanging as a credential, on a precision base with a small figure for scale and a soft coral wash, on warm paper.

ResearchOpen to read

11 June 2026

Agents Will Act for Us. Who Vouches for Them?

When agents start acting for us, trust moves from outputs to actors. Reading Michael Casey’s case for proof of control alongside our three-body picture of digital trust: where convenience quietly pulls authority toward a few global platforms, how validant.ai answers it, and a closed beta for scientific validation opening in Q3 2026.

Read

Swiss-style technical line drawing of three spheres of different sizes held in balance by interlacing elliptical orbits, with soft coral washes, on a white ground.

ResearchOpen to read

5 June 2026

Digital Trust Is an Orbit, Not a Pillar

Trust is not one more pillar to stack. It is the orbit three bodies trace together: the model, the person and the organisation. Why the three-body problem is the honest metaphor for trustworthy AI, and how to tell where you are in the orbit.

Read

Swiss-style technical line drawing of Lucerne’s covered wooden Chapel Bridge and octagonal Water Tower over the Reuss, with Mount Pilatus behind and a soft coral wash in the sky.

EventsOpen to read

30 May 2026

Two Views of One Decision: Trustworthy and Explainable AI in Practice at HSLU

Notes from a Lucerne specialists course on Trustworthy and Explainable AI, and what it confirms about the Validant.ai approach.

Read

Swiss-style technical line drawing of a single magnifying lens at the centre, identical lines fanning out from it to a row of small office buildings and a uniform crowd of figures, with one lone figure softly stained coral and set apart, in thin black lines on white.

StudiesOpen to read

27 May 2026

The Wrong Question, Asked at Scale

A landmark study of 4 million job applications shows how AI hiring tools hide their bias, why one rejection can become rejection everywhere, and why independent, position-level assessment is no longer optional.

Read

Swiss-style technical line drawing of five Trustworthy AI Circle participants smiling and holding up peace signs, rendered in thin black lines on white with soft coral accents.

EventsOpen to read

26 May 2026

No System Has Ever Been Fair

What four breakout sessions, one fairness tool demo, and 50+ years of collective experience taught us about fairness in AI, at the Trustworthy AI Circle.

Read

Still frame from the new validant.ai product demo videos: the Pulse audit view in motion.

UpdatesWatch preview

22 May 2026

Two new demo videos: see Pulse run an audit, and the Modules behind every verdict

We recorded the platform in motion — a full fairness audit start to finish, and the depth behind every result.

Watch & read

ResearchOpen to read

21 May 2026

Bias is the Foundation

Why every fairness claim begins with a bias diagnosis, and why skipping it breaks everything downstream.

Read

ProductOpen to read

2 May 2026

Inside Trust Signal: the weekly newsletter our AI team writes itself

Every Tuesday, an eight-agent AI team scouts, scores and drafts the week in AI trust and fairness. Here's how it works — and how to get it.

Read

FAQ

Frequently asked questions

What is validant.ai?: validant.ai builds systems for trust and agency. For any AI system that decides about people, it runs one workflow: assess the system, communicate the results to everyone who relies on them, and validate them as proof anyone can check. It covers three bodies: the model, the organisation behind it and the person it decides about. Use the platform yourself, or let us run the workflow with you.
Who signs what?: We sign the measurement: a seal attests the results our engine computed, by the stated method, on the data supplied. It is not a legal verdict and not a promise about any decision or its outcome. That conclusion is for your counsel, or for a partner auditor where a law asks for one.
How is it priced?: The platform comes in self-serve plans. When we run the workflow with you, with advice on what to change and how to govern it, it is part of an Enterprise subscription, scoped per AI system and priced per contract against the audit work it replaces, never on whether a system passes. A free first reading gets you started; pilots are paid.
What does validant.ai do for AI fairness?: It runs end-to-end AI fairness and explainability audits: detecting bias and proxy variables, mitigating disparities across data, models, LLMs and agents, explaining decisions with SHAP, LIME and counterfactuals, and monitoring fairness in production with auditor-grade, regulator-ready reports.
Who is validant.ai for?: Anyone who wants AI to stay fair, understandable and under human control. That includes risk, compliance, legal and ML teams proving an AI system is fair, explainable, trustworthy and accountable under the EU AI Act, NIST AI RMF and ISO/IEC 42001, and just as much the everyday person who wants to safeguard their own data and keep control of their agency. We deliberately keep it plain enough for everyone to use, not only specialists.
What is Verifiable Trust Infrastructure (VTI), and how does validant.ai relate to it?: VTI is an open specification, currently a working draft, developed in the Decentralized Trust Graph Working Group run jointly by Trust over IP and the Decentralized Identity Foundation. It defines how people, communities, organisations and the AI agents acting for them vouch for each other, delegate authority and revoke it without a central gatekeeper. validant.ai's founder is a contributor at both foundations and active in that working group, and validant.ai's seals, decision receipts and agent mandates are designed to plug into VTI as it matures.

Make AI fair, explainable and provable: make AI trustworthy.Make AIfair.

Fairness, assessed.Results, communicated.Proof, validated.

Assess

Communicate

Validate

One workflow, every decision about people.

Hiring and HR. Role by role.

Credit and lending

Insurance

Health and public sector

Generative AI and agents

Every AI decision touches three parties. We assure all three.

What your model actually does, with the receipts to prove it.

Proof that your organisation keeps its promises.

A record of every AI decision, held by the person it is about.

Where an algorithm already decides.

An open trust layer, for people and the AI that acts for them.

People, organisations and their AI

Trust Seal · Decision Receipt · Agent Mandate

Verifiable Trust Infrastructure

W3C DIDs and Verifiable Credentials

Checkable by anyone

People stay in control

Built in the open

One platform, three bodies, every claim auditable.

What you get.

Five commitments, built into the runtime.

Standards before features

Rigour a peer reviewer accepts

Compliance is the default

Auditable in production

Privacy and sovereignty by default

Three ways in, one workflow.

Do it yourself.

We run it with you.

Their judgement, our evidence.

You sign the conclusion. We supply the signed evidence.

The fairness library that powers the platform.

The AI Trust & Fairness Compendium

Signal.

You Cannot Audit What You Cannot Trace

Controlling Agents Triples Their Value. Proving It Is What Makes It Count.

The Preemptive Assurance Paradox: Who Guards the AI Guardrails?

Precision Is Not Proof: The Trap in Every AI Fairness Verdict

The Trust Problem Nobody Wants to Name

When a Model Becomes a Munition

Agents Will Act for Us. Who Vouches for Them?

Digital Trust Is an Orbit, Not a Pillar

Two Views of One Decision: Trustworthy and Explainable AI in Practice at HSLU

The Wrong Question, Asked at Scale

No System Has Ever Been Fair

Two new demo videos: see Pulse run an audit, and the Modules behind every verdict

Bias is the Foundation

Inside Trust Signal: the weekly newsletter our AI team writes itself

Get the week in AI trust and fairness.

Frequently asked questions

Make AI fair, explainable and provable: make AI trustworthy.