Asenion is an independent AI tester in the AI Verify Foundation's Global AI Assurance Pilot. We test real-world AI applications and agents against the rules that apply to them, using compliance-driven test data, local cultural context and domain experts. AI Assurance is our independent testing service. Asenion Verify is the product your own team uses to test AI before deployment.
Schedule a Call30 minutes with our AI assurance team. Bring the AI system you need to test.
The AI Verify Foundation pairs companies deploying AI with independent specialist testers to test real applications, not lab demos. Here is our work.
Mind Interview, a Taiwan-based HR AI company, scores candidate answers for first-round screening. We tested it for bias with synthetic candidate profiles that varied sex and race while holding answers constant, measured against the four-fifths impact-ratio rule used for NYC Local Law 144. We also reviewed the interview questions against US, UK and Canadian employment law, scanned for personal data, and stress-tested scoring quality against a benchmark model.
Read the case study →impress.ai and Asenion tested SAVOS, an agentic AI hiring assistant, identifying the risks of an autonomous recruitment agent and testing for them. The work was presented at the AI Verify Foundation's “Making AI Assurance Work in the Real World” session in January 2026 and at the AI Assurance Exchange during ATxSummit 2026.
Read the case study →At ATxSummit 2026, the AI Verify Foundation announced the AI Tester Accreditation Programme (AI TAP) and named Asenion among the testing firms taking part early. “We can prove that we are competent, there is independence, there is a consistent process and technique we follow.” — Fion Lee-Madan, Asenion
Read the recap →Beyond the AI Verify Foundation programme, we test and validate AI where trust matters most.
A US sponsor bank needed its tax loan model ready for an OCC fair lending exam. As an independent third-party validator, we ran a Level Two assurance validation against 35 controls aligned with ISO/IEC TR 24027 on bias in AI systems, automating qualitative and quantitative assessments for the bank's model validation and fair lending teams. After set-up, each quantitative re-validation took minutes, saving more than 100 hours.
Read the case study →One of the world's top banks faced a wall of overlapping AI security guidance: NIST, the OWASP Top 10 for LLM applications, MITRE ATLAS, the Cloud Security Alliance and ISO standards, each with its own language and scope. We made sense of it all, mapping the frameworks into one set of AI security controls, assessing the bank's AI against it, and showing exactly which requirements are covered and where the gaps are. One assessment, many frameworks, and findings the bank's security and risk teams can act on.
When AI reaches nearly 50 million mobile subscribers, a single failure is repeated millions of times. This telecom operator uses the Asenion AI Control System to create its own AI assessments and run pre-deployment red-teaming before new AI goes live. The testing identified vulnerabilities to fix before launch, and pinpointed where additional solutions from the AI ecosystem were needed to close the gaps, giving its risk and compliance teams evidence mapped to every control.
Led by Dr. Carrie Ye at the University of Alberta and funded by the Canadian Institutes of Health Research, ChatRheum answers rheumatology patients' questions while they wait to see a specialist. In healthcare, a wrong or careless answer can cause real harm, so we help test that its responses are accurate, empathetic, readable and safe for patients.
Read about ChatRheum →Asenion is a member of the Digital Health and Discovery Platform (DHDP). The Terry Fox Research Institute (TFRI) is the sole lead and governing body that manages and directs the DHDP. We support DHDP members in developing and adopting AI in healthcare with Asenion AI Assurance and the Asenion AI Control System, so health AI is tested, governed and evidenced from research through to clinical use.
Asenion on DHDP →Growthential guides organizations through AI adoption and business transformation. As Growthential's strategic alliance partner, Asenion brings independent AI assurance and compliance controls into its AI Strategic Suite, so clients build governance in from their first AI use case instead of adding it later.
Visit Growthential →We work with Canada's innovation programmes and leading advisory firms to help organizations build, test and certify AI responsibly.
Asenion is part of CENGN's $45 million Living Labs initiative, funded by the Government of Canada's Strategic Response Fund. Across eight Living Labs supporting over 100 Canadian startups and scaleups, we help companies prepare for ISO/IEC 42001 certification and independently test their AI for bias, privacy, security, accuracy and safety.
About CENGN Living Labs →ventureLAB supports innovative Canadian tech entrepreneurs with industry, academic and government partners. We offer ventureLAB companies support to get ready for ISO/IEC 42001 certification audits and independent AI assurance testing, so they can win enterprise customers with evidence, not promises.
About ventureLAB →EY Canada and Asenion (then Fairly AI) co-authored a white paper on responsible AI in high-risk sectors such as health care and financial services. It pairs EY's Confidence Index governance framework with Asenion's AI management system, which automates testing, monitoring and compliance scoring against standards such as ISO/IEC 42001, the NIST AI RMF and NYC Local Law 144: human-led governance, tech-empowered assurance.
Download the white paper (PDF) →A test report is only as good as the tester behind it. Here is what sets our assurance apart.
We have tested real deployments in the AI Verify Foundation's Pilot and Sandbox, we are part of its tester accreditation effort, and our results feed a public AI Trust & Safety Assurance Registry. Our reports stand up to regulators, customers and auditors.
In 2022, after a successful pilot with the British Standards Institution in preparation for EU AI Act audits, Fairly AI (now Asenion) became one of the first seven companies in the world included in the UK government's Portfolio of AI Assurance Techniques.
See the UK government listing →In 2023, before ISO/IEC 42001 became a published standard, we were the AI assurance tester in the Standards Council of Canada's pilot with the Responsible AI Institute and ATB Financial. We helped test and evaluate the controls for ATB's “Next Best Conversation” AI system, which received a provisional RAI Institute certification.
Read about the pilot →Generic benchmarks do not prove compliance. We source and build test datasets from the requirement itself, such as synthetic candidate profiles for bias audits, adversarial prompts for confidentiality and injection tests, and gold-standard answers for quality, so every result maps back to a control.
Fairness, harm and privacy mean different things across jurisdictions, languages and cultures. We design tests around the protected characteristics, norms and laws of each market you operate in, from North America and Europe to Asia-Pacific.
Our testers work with domain experts who know how AI is used and regulated in your sector, so we test the failure modes that matter to your business, not a generic checklist.
A repeatable process for chatbots, scoring models and AI agents, aligned with ISO/IEC 42001, the EU AI Act, NIST AI RMF and sector rules.
Map your use case to the regulations, standards and internal policies that govern it, and agree which risks to test: bias, privacy, accuracy, safety, security and agent behaviour.
Execute bias audits, red-teaming, privacy scans and quality benchmarks with test data built for your requirements, your markets and your industry.
Get findings mapped to controls, remediation guidance and a third-party attestation or validation you can share, with an optional listing in the AI Trust & Safety Assurance Registry.
In 30 minutes we'll scope your AI system, the rules it must meet and the tests that would prove it.
Schedule a CallNo preparation needed.
