The EU Artificial Intelligence (AI) Act Compliance

Module 1: Introduction to the EU AI Act (30 minutes)

• AI compliance training begins with an overview of the legislative framework and objectives of the EU AI Act

• Key definitions: AI system, provider, deployer, high-risk AI, general-purpose AI (GPAI)

• Scope and applicability: Who needs to comply?

• Timelines for implementation

Module 2: Risk Classification of AI Systems (45 minutes)

• Four-tier risk classification: Unacceptable risk (prohibited AI systems)

• High-risk AI systems and regulatory obligations

• Limited risk AI systems and transparency requirements

• Minimal risk AI systems (unregulated but encouraged to follow best practices)

Case studies of AI system classification

Module 3: Prohibited AI Practices (45 minutes)

This module explores the prohibited practices set out in Article 5 of the EU AI Act, examining high-risk AI applications and the legal consequences of non-compliance.

• Overview of Article 5 of the EU AI Act

• Prohibited AI applications: Subliminal/manipulative AI causing significant harm

• Exploitation of vulnerable populations

• Biometric categorisation of sensitive attributes

• Social scoring systems

• Predictive policing based on profiling or personal traits

• Mass facial recognition via untargeted scraping

• Emotion recognition in workplaces and education (exceptions apply)

• Real-time biometric identification in public spaces (limited exceptions)

Module 3 of AI Act training concludes with the legal consequences of non-compliance

Module 4: Compliance Requirements for High-Risk AI Systems (1 hour)

This module examines the regulatory obligations for both providers and users of high-risk AI systems, focusing on lifecycle risk management, oversight, and practical implementation requirements.

• Identifying high-risk AI use cases (Annex I & III)

• The obligations for providers (developers) of high-risk AI: Risk management systems throughout the AI lifecycle

• Data governance and dataset quality requirements

• Technical documentation and compliance demonstration

• Logging and record-keeping

• Human oversight mechanisms

• Accuracy, robustness, and cybersecurity requirements

• Quality management systems

The obligations for deployers (users) of high-risk AI:

• Monitoring and reporting obligations

• Ensuring human oversight

• Data protection and transparency duties conclude module 4 of AI compliance training

Case study: Implementation of high-risk AI compliance requirements

Module 5: General-Purpose AI (GPAI) and Systemic Risk (45 minutes)

• The definition and obligations for GPAI providers

• Compliance requirements: Technical documentation

• Copyright Directive compliance

• Summary of training data

• Cooperation with high-risk AI system providers

Systemic risk criteria and obligations for high-impact GPAI:

• Model evaluation and adversarial testing

• Risk assessment and mitigation

• Serious incident reporting

• Cybersecurity protections

Role of voluntary codes of practice and harmonised standards

Module 6: Governance, Enforcement, and Penalties (30 minutes)

• Role of the AI Office and national supervisory authorities

• Mechanisms for compliance monitoring and enforcement

• Complaint mechanisms for downstream providers

• Penalties for non-compliance and risk mitigation strategies

Module 7: Practical Compliance Strategies for Organisations (45 minutes)

Participants gain practical strategies to embed compliance and prepare for regulatory review in this section of our AI compliance course.

• Developing an AI compliance framework

• Conducting AI impact assessments and risk classification

• Integrating legal, ethical, and technical compliance practices

• Documentation best practices for audits and regulatory reviews

• Case study: Best practices from leading AI companies

Q&A and Closing Discussion (30 minutes)

• Open floor for participant questions

• Final thoughts on navigating compliance challenges

• Recommended next steps for organisations

Training Objectives

AI compliance training ensures participants gain the knowledge and skills to:

• Understand the legal and regulatory framework for AI in the European Union

• Recognise changes and new obligations imposed by the EU AI Act

• Identify and assess the risks associated with AI system development and deployment

• Implement effective governance and compliance strategies for AI systems

• Ensure adherence to AI transparency and accountability requirements

• Navigate the ethical implications of AI

• Maintain up-to-date knowledge of the latest developments and best practices in AI

Training Course Summary

AI Act Training at Redcliffe provides participants with a comprehensive understanding of the compliance framework under the EU’s AI regulation. Sessions are delivered by an experienced specialist with a long track record in the global financial markets and specialisation in data privacy compliance implementation.

Whether you are a provider, developer, deployer, manager, adviser, regulator or investor, this programme provides the knowledge needed to prepare for the evolving regulatory landscape and support. This AI compliance course is delivered at an agreed location.

Case Study: You will use a variety of hypothetical and real case studies and exercises (the latter, based on publicly available information).