Modulos AI Governance Platform
Modulos AI Governance Platform is a Swiss GRC workspace for regulated organisations. It maps one shared control library across 26 frameworks, quantifies AI risk in money, and lets human-supervised AI agents assemble the audit evidence.
What is Modulos AI Governance Platform?
Modulos AI Governance Platform is the product of Modulos AG, a Zurich company spun out of ETH Zurich and entered in the Swiss commercial register on 18 June 2018. It is a governance, risk and compliance workspace built for one job: proving that an organisation's AI systems meet the rules that apply to them.
The platform is organised in four pillars. Governance holds the inventory: project and organisation dashboards, AI lifecycle stages, ownership workflows and complete audit trails. Risk prices AI exposure in money rather than in a colour-coded matrix, using Fermi estimation and Monte Carlo methods, with appetite limits, a risk taxonomy and risk-to-control mapping. Compliance rests on a shared library of 26 frameworks and 958 controls; 454 of those controls serve two or more frameworks and the most reused serve ten, so evidence attaches to the control rather than to the framework and is collected once. Agents covers four human-in-the-loop assistants: Scout answers compliance questions and hunts evidence in connected GitHub, Confluence and Google Drive repositories with source citations; a control assessment agent turns a two-to-four-hour manual review into roughly five minutes plus human review; an evidence agent produces metadata, summaries and mapping suggestions from uploaded documents; a risk agent audits its own reasoning before returning a figure. Every recommendation still requires a human approval.
Coverage spans the EU AI Act, ISO/IEC 42001, ISO/IEC 27001, ISO/IEC 27701, NIST AI RMF, GDPR, DORA, NIS2, MAS FEAT, Singapore's framework for agentic AI, UAE AI Ethics and the OWASP top tens for LLMs and agents. Runtime tests close the loop, checking thresholds on live metrics from Prometheus, Datadog or CloudWatch and running natural-language compliance checks across connected systems.
Customer models and training data stay where they are. Modulos ships as SaaS or in a private cloud or VPC, is SOC 2 Type 2 certified, claims first ISO/IEC 42001 product conformity, and offers an interface in English, German, French, Italian and Spanish.
What it does
- Register every AI system with an owner, a lifecycle stage and an approval status
- Classify a system under the EU AI Act and track the evidence requirement by requirement
- Implement a control once and have it count across every framework that maps it
- Put a monetary figure on AI risk and set risk appetite limits
- Let AI agents assess controls, triage evidence and draft recommendations for human approval
- Run scheduled runtime tests on production metrics and link the results back to controls
- Export an audit-ready record for a regulator, an auditor or a customer questionnaire
When to use Modulos AI Governance Platform / When not to
A quick filter to help you decide if Modulos AI Governance Platform is the right fit.
When to use Modulos AI Governance Platform
- Compliance and risk officers who must show a regulator, an auditor or a customer how their organisation governs AI
- Teams preparing an ISO/IEC 42001 certification and needing gap analysis, control ownership and audit trails in one place
- Organisations classifying AI systems under the EU AI Act and tracking evidence requirement by requirement
- Enterprises running several frameworks at once who want a control implemented once to count everywhere it applies
- Regulated sectors named by the vendor: financial services, insurance, transport, utilities, mobility, telecommunications and defence
When not to use Modulos AI Governance Platform
- Teams looking for a data governance platform: the vendor states plainly that this is not what Modulos does
- Anyone needing an observability or MLOps stack, which the vendor says it does not replace
- Buyers who want to sign up, see a price and start on their own: every path leads to a demo request form
- Mobile-first users, since there is no iOS or Android application
- Free-tier users who need vendor support, because technical support is excluded from the Freemium terms
How to use Modulos AI Governance Platform
A typical end-to-end flow, from setup to results.
- Submit the demo request form, which the vendor says takes about thirty seconds; there is no self-service sign-up
- Receive a confirmation within twenty-four hours
- Take a thirty-minute intro call covering your AI portfolio, governance maturity and compliance priorities
- Sit through a platform demo built around the workflows, controls and evidence needs you named
- Alternatively, request the Free Starter Plan, which needs no credit card and is granted after approval with onboarding support
- Sign in to the workspace at app.modulos.ai
- Register your AI systems and assign owners and lifecycle stages
- Connect your sources: GitHub, Bitbucket, AWS, Azure, Google Drive, Confluence, Jira, MLflow, Jupyter, Snowflake or Copilot Studio
- Wire operational metrics in through the Python SDK, the REST API, webhooks, a direct database connection or file imports
- Select the frameworks that apply, let the agents draft assessments, review them, and schedule the runtime tests
Pros & Cons
Pros
- One control implemented once counts across every framework that maps it, and the vendor publishes the numbers behind the claim
- Risk is expressed in money rather than in a qualitative matrix, which is what boards and auditors ask for
- Strong certification posture: SOC 2 Type 2 plus a claimed first ISO/IEC 42001 product conformity
- Customer models and training data never leave the customer's environment
- Private cloud and VPC deployment give full data sovereignty, with a dedicated SLA
- Published Python SDK and REST API make it possible to push metrics straight from a CI/CD pipeline
- Interface available in English, German, French, Italian and Spanish on every deployment mode
Cons
- No public pricing at all: no pricing page exists anywhere in the 210-URL sitemap, and every route leads to a demo request
- Nothing is self-service, not even the free plan, which needs a form and an approval
- The Freemium terms grant Modulos an irrevocable, perpetual right to train on anonymised customer data with no documented opt-out
- The free plan excludes technical support entirely
- No EU Article 27 representative is named, which is unusual for a Swiss vendor selling into the European Union
- No data processing agreement is published: you have to write to privacy@modulos.ai to get one
- Several published surfaces are stale or down, including an outdated FAQ page and an MCP server the vendor itself flags as disabled
Pricing & Plans
A permanent free plan exists. The Free Starter Plan, launched in January 2025, covers one user, one project and up to one hundred AI agent workflows per month across the public EU AI Act, ISO 42001 and NIST AI RMF frameworks; no credit card is required, access is granted after approval, and technical support is not included. No paid price is published. There is no pricing page anywhere on the site, and the terms and conditions refer instead to an individually negotiated Commercial Offer. Where that offer provides for usage-based billing, fees combine a Base Quota of Consumption Units per measurement period with usage fees invoiced in arrears for anything beyond it. A Pricing Schedule setting the conversion rates is described in the contract as published on the website, but it could not be found there. Because no amount is public, the lowest paid entry point cannot be stated.
- Free Starter Plan — no charge
- one user
- one project
- up to 100 AI agent workflows per month
- public EU AI Act
- ISO 42001 and NIST AI RMF frameworks
- no credit card
- access granted after approval
- Commercial Offer — the paid tier
- scope
- term and price are negotiated individually and governed by the Terms and Conditions
- may include usage-based billing in Consumption Units
- no named tier and no amount is published
- SaaS deployment — no infrastructure to manage
- automatic updates and scaling
- SOC 2 Type II
- interface in English
- German
- French
- Italian and Spanish
- available through cloud marketplaces
- Private Cloud or VPC deployment — full data sovereignty
- custom network configuration
- dedicated support and SLA
- interface in English
- German
- French
- Italian and Spanish
Data, GDPR & hosting
A consolidated view of how Modulos AI Governance Platform handles your data.
GDPR overview
The privacy policy is concrete rather than decorative. Version 4.0, effective February 2026, it covers both the website and the app.modulos.ai platform, names Modulos AG in Zurich as controller and routes requests to privacy@modulos.ai. It cites Article 6(1)(f) as the lawful basis for its CRM processing, sets out access, rectification, erasure, portability, restriction and objection rights, and relies on Standard Contractual Clauses and, where applicable, the EU-U.S. Data Privacy Framework for transfers. As a Swiss company Modulos is also subject to the revised nFADP, and states that where both regimes apply it applies the higher standard. Thirteen sub-processors are published with purpose and country. Two things are missing: no EU representative is designated under Article 27, and no data protection officer is named.
Who owns the data?
Modulos AG owns the platform, its software and its trademarks; the customer keeps its own models and training data, which never leave its environment. What differs is the licence the customer grants back. Under the paid Terms and Conditions, Modulos may collect metrics, analytics and usage statistics to run the service and to improve the platform, and that second purpose explicitly excludes personal data. Under the Freemium terms the licensee grants Modulos a worldwide, irrevocable, perpetual, royalty-free right to use anonymised licensee data and metadata to train the platform and its AI/ML models, and that grant survives the end of the licence. Inactive accounts are flagged after thirty days and deleted after ninety.
Reuse rights
The customer keeps full use of its own material: the platform connects to existing repositories and tools rather than ingesting datasets or models, so nothing has to be handed over and nothing has to be asked back. Evidence, assessments and audit exports produced inside the workspace belong to the customer's own record and can be taken out through audit-ready exports, the REST API or the Python SDK. What the customer may not do is turn the platform itself around: the Terms and Conditions forbid sublicensing, reselling, copying, modifying, creating derivative works from, reverse-engineering or decompiling it, and forbid letting a third party do so. Pricing terms and the commercial details of the agreement are treated as confidential information.
Data retention & training
Hosting summary
The privacy policy states that Modulos stores, uses and processes personal data in Switzerland or within the European Economic Area, and that its proprietary CRM at crm.modulos.ai runs on third-party cloud infrastructure located in the European Union. Transfers out of Switzerland rely on adequacy decisions, Standard Contractual Clauses recognised under Swiss law or other lawful mechanisms; transfers to the United States rely on Standard Contractual Clauses and, where applicable, the EU-U.S. Data Privacy Framework. The product pages describe something broader: data residency can be chosen across the EU, the United States, the United Arab Emirates, Singapore and further regions, or kept entirely on the customer's own infrastructure in a private cloud or VPC deployment. Data is encrypted with AES-256 at rest and TLS 1.3 in transit. Thirteen sub-processors are published with their country: Germany, Switzerland, the Netherlands and, for nine of them, the United States.
Things to keep in mind
Risks and trade-offs to weigh before adopting Modulos AI Governance Platform.
- Under the free plan you sign away an irrevocable, perpetual right for Modulos to train its models on your anonymised data and metadata, and that grant outlives the licence with no opt-out documented anywhere
- The site tells you what the rules are and what evidence to keep, but it cannot certify you: an auditor or a notified body still has to, and a platform can make an unfinished programme feel finished
- The vendor publishes contradictory statements about where data lives, placing processing in Switzerland or the EEA in the privacy policy while advertising EU, US, UAE and Singapore residency on the product page; nine of its thirteen sub-processors are American
- No EU Article 27 representative and no published data processing agreement, which is precisely the kind of gap a governance tool is bought to prevent
- Agents draft assessments and recommendations that a human is meant to approve, and approval fatigue is a real risk once the volume rises: the readiness score is a suggestion, not a verdict
- Parts of the published surface are stale or broken, including an FAQ page that still expects the EU AI Act to be published in April 2024, a dead YouTube link and a public MCP endpoint the vendor flags as disabled
- Pricing is invisible until you are in a sales conversation, and the contract lets Modulos amend its usage-fee schedule with effect immediately after notice
Setup & Integrations
Technical difficulty
Getting in is a sales process, not a technical one: a form, a confirmation within a day, an intro call and a demo, with the free plan granted after approval and onboarding support. The technical effort comes afterwards and scales with ambition. Connecting repositories and document stores such as GitHub, Confluence or Google Drive is straightforward. Feeding runtime tests is engineering work: the Python SDK, the REST API, webhooks or a direct database connection have to be wired into a CI/CD pipeline. Private cloud or VPC deployment adds network configuration, and eight access roles have to be assigned.
Deployment
Integrations
Supported languages
Behind Modulos AI Governance Platform
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Modulos AI Governance Platform.
Frequently asked questions
Is Modulos a data governance platform?
Do I have to send Modulos my data?
What does it cost?
Which frameworks does the platform cover?
Is there an API?
Does Modulos train its models on my data?
Where is my data hosted?
What languages does the interface support?
What security certifications does Modulos hold?
How do I get a data processing agreement?
Should you pick Modulos AI Governance Platform?
Modulos has picked a narrow position and defends it well. It is not a data governance suite, not an observability stack and not an MLOps tool, and the vendor says so on its own pages. What it sells is the record: an inventory of AI systems, a control library shared across twenty-six frameworks, evidence attached to controls rather than to checklists, and a monetary figure on risk that a board or an auditor can actually use.
The credibility markers are unusually solid for a company of this size. SOC 2 Type 2, a claimed first ISO/IEC 42001 product conformity, a place in the inaugural Gartner Magic Quadrant for the category, named institutional customers including a European financial supervisor, and CHF 16.4 million raised behind a 2018 ETH Zurich spin-off.
The reservations are real and worth weighing. Pricing is entirely opaque: there is no pricing page at all, and even the free tier requires a form and an approval, so no buyer can size the commitment before entering a sales conversation. The Freemium contract grants Modulos an irrevocable and perpetual right to train its models on anonymised customer data, with no opt-out documented anywhere, which sits oddly with a product whose whole argument is governance discipline. And for a Swiss vendor selling into the European Union, the absence of a named Article 27 representative and of any published data processing agreement is a gap a compliance buyer will notice first.
For an organisation that already knows it must certify against ISO/IEC 42001 or evidence EU AI Act obligations across several frameworks, Modulos is a serious candidate. For anyone still comparing on price or wanting to try before talking, the door is closed.
- Choosing a selection results in a full page refresh.
- Opens in a new window.