Intertek AI² enables organisations to power ahead with smarter, safer, trusted AI solutions
AI is rapidly accelerating in all parts of society, quickly changing the way the world works and triggering significant risks for governments, corporates, consumers and employees.
Intertek AI2 is the world’s first independent end-to-end assurance programme that will give our clients a competitive advantage with smarter, safer, trusted AI solutions enabling them to power ahead.
Intertek AI2 is a comprehensive portfolio of Consulting, Training, and Assurance solutions that supports your product’s entire AI lifecycle—from ideation to deployment, and beyond.
Whether you're developing a smart device, incorporating machine learning into existing systems, or entering new markets with AI features, Intertek delivers risk-based quality and safety solutions tailored to your needs. Our approach is flexible, scalable, and aligned with global best practices and regulatory expectations.
Intertek’s comprehensive AI service framework is built around four core pillars of AI assurance: governance, transparency, security, and safety. These frameworks combine the state of the art in international AI standards, industry best practices, and guidelines. Modules are designed to evaluate how AI systems are managed, how decisions can be explained, how systems are protected, and how risks are controlled. The result is a structured, auditable approach to AI assurance that adapts to your sector, risk profile, and regulatory obligations.
Governed AI
Governed AI establishes risk and quality management frameworks, AI governance structures, regulatory compliance strategies, and oversight mechanisms ensuring accountability and adherence to evolving requirements including EU AI Act obligations.
We deliver expert-led ISO 42001 training to help your team understand and implement AI management systems that align with international best practices. The training helps balance innovation with governance by addressing AI-specific risks like ethics, transparency, and continuous learning. Participants gain a clear understanding of why AI governance matters, how to identify and manage AI-related risks, and how to build an AI Management System (AIMS) aligned with ISO/IEC 42001. The training also shows how ISO 42001 integrates with existing standards to create a unified, trustworthy approach to managing AI.
We provide auditing and certification of the AI management system per ISO/IEC 42001 requirements.
We strengthen AI development practices by assessing organizational controls with ISO/IEC 42001. Our work supports structured quality objectives, lifecycle documentation, and improvement loops specific to AI systems.
We apply ISO/IEC 23894 and the NIST AI Risk Management Framework to help identify, assess, and treat AI-specific risks across the system lifecycle. Our approach embeds risk thinking into both technical and organizational decision-making.
We assess governance structures aligned with ISO/IEC 42001 and the requirements of the EU AI Act. From roles and responsibilities to documentation and audits, our services help organizations operationalize control over AI systems at scale.
Secure AI
Secure AI delivers cybersecurity tailored to AI systems, red teaming exercises identifying vulnerabilities and failure modes, threat monitoring and incident response planning, and security architecture guidance addressing unique AI vulnerabilities.
Combining Functional Safety, Cybersecurity, and AI, we help ensure AI behaves as intended, even in adversarial or failure conditions. Our work supports both technical resilience and regulatory defensibility.
We identify vulnerabilities and cyber risks in AI systems using principles from ISO/IEC 23894, OWASP Machine Learning Security Top Ten, and OWASP Top 10 for Large Language Model Applications. This includes analyzing risks like data poisoning, evasion attacks, and model extraction.
We evaluate AI systems for cybersecurity risks including cybersecurity risks related to supporting infrastructure and systems, AI applications, and AI models. Our work focuses on securing the entire AI systems, including vulnerability assessments, penetration testing, and AI red teaming.
We support clients validate the use of AI/ML in the application of digital twin simulations to assess system behavior. This enables early identification of risk scenarios and ensures digital twin technology that leverage AI can confidently be used speed up compliance.
Transparent AI
Transparent AI develops technical documentation meeting regulatory standards, implements appropriate explainability levels for different applications, and creates communication strategies making AI behavior understandable to diverse stakeholders.
Our pre-assessments identify gaps against current and emerging AI regulations and standards, including the EU AI Act and ISO/IEC standards. You’ll get a clear action plan to accelerate compliance and reduce costly surprises during formal evaluations.
Aligned with ISO/IEC 42001 and the requirements under the EU AI Act, we review how human oversight is defined and implemented. This includes evaluating intervention points, escalation paths, and decision accountability for high-risk AI systems.
We assess the quality of datasets used in AI systems based on the ISO/IEC 5259 series, which defines key dimensions such as accuracy, completeness, timeliness, and representativeness. Our reviews focus on how data quality affects model performance, safety, and compliance, particularly in high-risk applications. We help ensure structured, auditable processes for managing and improving data throughout the AI lifecycle.
We detect and evaluate bias using the principles in ISO/IEC TR 24027 and ISO/IEC TS 12791. Our reviews help ensure fairness in datasets and model behavior and guide mitigation actions when disparities are found.
We assess documentation to evaluate explainability techniques, including addressing data governance, human oversight to ensure model clarity. This ensures your AI systems are transparent, traceable, and aligned with upcoming regulatory transparency demands.
Safe AI
Safe AI provides comprehensive testing and validation using AI-specific methodologies, data quality assessment and improvement, independent performance verification, and bias detection and mitigation across diverse populations and use cases.
We assess performance, focusing on performance claims, stability, generalization, drift, and resilience. Our work ensures models perform reliably across deployment conditions, not just in the lab.
Using ISO/IEC 24029-1:2021 guidance, we test AI systems for resilience under stress, data noise, and edge cases. These assessments evaluate how well your models perform in real-world, uncertain, or adversarial conditions to ensure they’re stable and trustworthy.
We assess how AI-driven components interact with safety functions in systems such as robotics and machinery, aligning with relevant functional safety standards such as IEC 61508, ISO 13849 and ISO 26262. Our focus is on demonstrating that AI behavior is predictable, controllable, and safe. We help ensure AI doesn’t compromise functional safety, particularly in systems involving dynamic decision-making, human interaction, or autonomous motion.
We evaluate how understandable and transparent your AI system is, from model outputs to lifecycle documentation, ensuring that We help make black-box systems are explainable to users, auditors, and regulators.
We support clients to develop and perform robust test strategies for AI that help verify and validate model behavior, outcomes, and support traceability from training to deployment.
Why Choose Intertek AI2?
- End-to-End Support: From concept to compliance, we guide you through regulatory, performance, and risk challenges.
- Global Reach: With over 1,000 laboratories and offices in more than 100 countries, we bring deep local knowledge to global market entry.
- TQA Expertise: Our Science-based Total Quality Assurance (TQA) model gives clients a competitive edge with early insights into current and emerging regulations.
- Cross-Industry Experience: Drawing on more than 130 years of safety and quality leadership, we’re uniquely positioned to help organizations safely and confidently adopt AI.
Trusted AI Begins Here
Intertek AI2 services help you validate that your AI systems are transparent, explainable, accountable, and ethical. Our solutions reduce time to market, minimize risk, and ensure lasting confidence among users, investors, and regulators alike.
Whether you're designing, deploying, or maintaining an AI-enabled product, Intertek ensures that safety, performance, and compliance are built in, not just bolted on.
Start your journey to responsible AI today.
Contact us to learn how Intertek AI2 can help you build safer, smarter, and more trusted AI technologies.
AI Compliance in Industrial Machinery
Minimize risk and meet regulatory requirements for AI-Enabled Industrial Machinery
AI Enabled Medical Devices
Tailored solutions for Medical Devices that address AI data quality, algorithm transparency, and compliance with global regulatory frameworks.
Intertek AI² - AI Assured: Smarter, Safer, Trusted
