AI is moving faster than compliance can keep up.
Our teams have been building the control system for AI since 2016. Asenion was formed in 2025.
Asenion exists to solve a problem that gets harder every time an organization puts another AI model, copilot or agent into production: how do you turn what an organization is required to do, and what it intends to do, into controls that AI systems can actually follow, verify and prove?
AI compliance is not simply a regulatory problem. It sits at the intersection of law, policy, risk, technology, security and business intent. That is why Asenion was built differently.
That is the foundation of the Asenion AI Control System.
Asenion turns the language of compliance into operational controls.
AI compliance is not fundamentally a document problem. Regulations, standards and policies are written in the language of humans. AI systems operate in the language of machines. The difficult, and increasingly valuable problem is the translation between the two.
Asenion transforms authoritative requirements into context-aware controls that can be implemented, verified and evidenced, creating a continuous connection between what your organization is required to do and what your AI systems actually do.
From the language of compliance to the language of machines.
Know your AI is under control, and be able to prove it.
Organizations come to Asenion when AI is moving into production faster than their compliance, risk and security teams can keep up. Here is what they get.
Controls that fit your business
We turn regulations, standards and your own policies into context-aware controls for your specific AI use cases, jurisdictions and risk appetite. Not a generic checklist.
Proof, not paperwork
Every control is assessed, verified and witnessed, producing tamper-resistant evidence that auditors, regulators, customers and your board can examine.
Expert-built, not just generated
Anyone can ask an LLM to write a control. Ours are built and reviewed by AI governance, legal and model risk experts, and refined across more than 300 enterprise AI assessments.
One system, the whole lifecycle
Assess AI use cases before they launch, verify models and agents before deployment, and enforce controls at runtime, with every result tied back to the same requirement.
Faster to compliance
Ready-made Policy Packs for the EU AI Act, ISO/IEC 42001, NIST AI RMF, U.S. state laws and more mean your teams start from expert controls, not a blank page.
Works with what you already use
Connect to your GRC, ITSM, cloud ML and agent platforms, so AI compliance fits into existing workflows instead of becoming another silo.
Generating a control is easy. Knowing it is fit for purpose is the hard part.
That is the problem Asenion has spent a decade solving, for organizations where AI decisions have real consequences.
Built for the organizations that cannot afford to guess.
Asenion is designed for organizations where AI decisions have material consequences and where evidence matters.
These organizations do not simply need to know what their policies say. They need to know whether their AI systems are actually operating within them, and be able to demonstrate that to the people who matter.
From regulatory language to machine-executable controls.
Most organizations do not struggle to find another regulation, framework or policy document. They struggle to answer one question: what does this requirement actually mean for us?
A regulation describes an obligation. A business has a context. An organization has an intent. A control must produce a specific outcome. And someone eventually has to prove that the control works. Asenion connects those layers.
Context-aware, intent-aware, machine-executable controls.
Our technology transforms regulatory, organizational and security requirements into controls that can be assessed, verified and witnessed across the AI lifecycle. That is the problem we have spent years learning how to solve.
Our technology is built around the control, not the document.
Most AI governance begins with documents. Asenion begins with the control.
A regulation may contain thousands of words. A policy may span dozens of pages. A standard may contain hundreds of requirements. The operational question is simpler.
Asenion’s technology is designed to answer that question systematically. Our platform connects regulatory and policy requirements to machine-readable controls, AI assessments, adversarial tests, runtime governance and evidence.
This architecture allows the same underlying compliance intelligence to support multiple modes of governance, from human assessment to automated testing to runtime controls. It also creates a foundation that can evolve as regulations change, AI architectures change and new forms of AI emerge.
What must the organization actually control, verify and prove?
Proprietary technology. Accumulated expertise. Compounding data.
Our defensibility is not based on a library of documents. It is built across multiple layers of technology and accumulated knowledge.
Compliance ontology
A structured model for representing requirements, controls, applicability and relationships across regulatory and standards sources.
Compliance-to-control translation
Technology and methodology for transforming natural-language requirements into operational controls.
Control and testing architecture
A common representation connecting controls to assessments, executable tests and outcomes.
AI evaluation technology
Methods for testing predictive, generative and agentic AI for security, privacy, fairness, hallucination and other risks.
Evidence architecture
A traceable record connecting requirements, controls, tests, results and evidence.
Accumulated compliance intelligence
Years of mappings, interpretations, test methodologies, edge cases and implementation knowledge developed through work with enterprise AI systems and global standards.
Together, these layers form the foundation of the Asenion AI Control System.
Built from experience, not theory.
Asenion was formed in 2025 through the acquisition of Swedish AI governance company anch.AI by Canadian AI governance company Fairly AI. It brought together two teams that had been working on AI risk before the market had settled on the term “AI governance”: one from North America and one from Europe.
Fairly AI pioneered compliance-focused AI testing and assurance tooling. anch.AI brought nearly a decade of methodology-driven experience assessing AI use cases and organizational risk. Together, our team has more than 30 years of experience in AI governance across academia and industry, and has assessed more than 300 enterprise AI models for ethical, legal and regulatory risk.
But our experience is not just about understanding AI risk. It is about operationalizing it. We have worked with organizations where AI must satisfy regulators, auditors, security teams, model risk teams, product leaders and business owners at the same time. We have seen where frameworks break down in practice, and what it takes to turn principles into controls, controls into verification, and verification into evidence. That experience is built into Asenion.
Combined AI governance experience across academia and industry
Enterprise AI systems assessed for ethical, legal and regulatory risk
Research and methodology development across Canada and Europe
Certified AI Control System
Two origins. One technology platform.
Fairly AI
Fairly AI emerged from financial services and the discipline of model risk management. The company grew out of Accenture’s FinTech Innovation Lab and applied the rigor of established model-risk practices to emerging AI systems.
Fairly AI developed independent AI testing capabilities for areas including fairness, privacy and security, and built technology aligned with emerging AI regulations and standards.
anch.AI
anch.AI emerged from the AI Sustainability Center in Stockholm, founded to research the ethical and societal implications of AI. Its work focused on AI risk methodology, organizational governance and the practical application of emerging European AI regulation.
With the EU AI Act central to its methodology, anch.AI worked with public and private organizations across the Nordic region.
Asenion
In 2025, these capabilities came together.
Research + regulation + model risk + verification + software + enterprise execution.
One platform.
Built by people who understand the problem from different sides.
Our team brings together disciplines that rarely exist under one roof.
AI research
Machine learning, frontier models and applied AI research.
Model risk
Financial model risk management and quantitative governance.
Regulatory and policy expertise
AI regulation, standards and organizational governance.
Software engineering
Enterprise-grade systems for operationalizing AI controls.
Product
Turning complex regulatory and technical requirements into usable enterprise software.
Enterprise growth
Experience building partnerships and technology businesses in highly regulated markets.
That combination matters. AI governance is not only a regulatory problem. It is not only an AI problem. And it is not only a software problem.
It sits at the intersection of all three.
We have lived through every generation of AI governance.
The next generation of AI governance is a control system.
AI is moving faster than traditional governance processes can keep up. Static policies cannot continuously govern dynamic AI systems. Periodic assessments cannot fully explain what an agent did at runtime. And documentation assembled after the fact is fundamentally different from evidence created as systems operate.
The hard part of AI compliance is not generating another policy. It is knowing what that policy needs to mean. Asenion combines regulatory and organizational intelligence with machine learning, control engineering and AI assurance to make that meaning operational.
This is Asenion.
Leadership
AI governance researchers, model risk practitioners and product builders from Canada and Europe.
David Van Bruwaene
Co-Founder & Chief Executive Officer
Machine learning expert and serial entrepreneur with academic roots in cognitive science and philosophy at Cornell University. David leads Asenion's vision for translating human and organizational intent into controls machines can operationalize.
Anna Felländer
Co-Founder & President, Asenion Europe
Former Chief Economist and pioneer in AI governance since 2016. Anna brings the institutional perspective: how regulation, markets, risk appetite and organizational decision-making shape what AI should, and should not do.
Fion Lee-Madan
Co-Founder & Chief Product Officer
Technology and digital transformation leader with more than two decades of experience. Fion leads the product strategy that brings Asenion's compliance intelligence, controls and assurance capabilities into practical enterprise workflows.
David Schifman
Chief Revenue Officer
Enterprise sales and partnerships leader from Apple, Meta and BlackBerry.
Samantha Chan
Chief Technology Officer
Former IBM research leader and technology architect with experience spanning enterprise software, healthcare technology and open-source systems.
Ahmed Orabi
Co-Chief Science Officer
AI researcher and frontier-model developer with experience at IBM and the University of Ottawa. Together with Mahmound known as "The Brothers".
Mahmoud Orabi
Co-Chief Science Officer
AI researcher and frontier-model developer with experience at IBM and the University of Ottawa. Together with Ahmed known as "The Brothers".
See what controlled, provable AI looks like in your organization.
Talk to our experts about your AI use cases, the regulations you face and the evidence you need. Kitchener-Waterloo, Canada · Stockholm, Sweden