Who is affected by the AI Act?
Discover the different sectors and actors concerned by the European AI regulation and their specific obligations.
The European Artificial Intelligence Regulation (EU AI Act 2024/1689) entered into force on August 1, 2024. Depending on your industry and the type of AI systems you use or develop, your obligations vary considerably. Discover below the concrete implications for your sector.
SaaS and AI startups
Whether you develop chatbots, recommendation engines or generative AI tools, you are concerned by the AI Act. Your obligations vary according to your risk classification.
As a SaaS publisher, you are considered an AI system "provider" under the AI Act. This implies specific obligations regarding technical documentation, risk management, and transparency towards your users. If your solution integrates foundation models (GPT, Claude, etc.), you must also comply with obligations for general-purpose AI systems.
Example systems:
- Customer service chatbots
- Predictive analytics tools
- Content generators (text, image)
- Lead scoring systems
Your obligations under the AI Act
Mandatory technical documentation, conformity assessment, CE marking for high-risk systems, and serious incident notification. Chatbots must inform users they are interacting with an AI (Article 50).
HR and Recruitment
AI systems used in recruitment and human resource management are classified as high-risk under Annex III of the AI Act. They require strict compliance.
Annex III of the regulation explicitly classifies AI systems used for recruitment and human resources management in the "High Risk" category. This classification applies when AI influences decisions having a significant impact on people's professional lives: candidate selection, skills assessment, promotion or dismissal.
AI systems for recruitment, candidate selection or employee evaluation are subject to enhanced obligations (technical documentation, human oversight, conformity assessment).
Concerned systems:
- Automated CV sorting
- Video interview analysis
- Performance evaluation systems
- Emotion analysis and lie detection (prohibited in certain cases)
Specific High-Risk obligations
Risk management system, high-quality data governance, comprehensive technical documentation, transparency and candidate information, effective human oversight, and conformity assessment by a notified body in certain cases.
E-commerce and Retail
E-commerce platforms using AI for recommendations, customer service or pricing have transparency obligations under Article 50.
For e-commerce players, most AI systems fall under limited risk, with obligations mainly related to transparency. However, beware: if your dynamic pricing system creates discrimination or if your chatbot can be mistaken for a human, additional obligations apply.
Concerned systems:
- Product recommendation engines
- Chatbots and virtual assistants
- Dynamic pricing systems
- Intelligent search
- Virtual assistants and AI try-ons
Note: If your site uses a chatbot or generative AI, you must inform users that they are interacting with an AI (Article 50).
Transparency obligations
Clearly inform users when they interact with a chatbot or AI-based recommendation system. If you use deepfakes or AI-generated content, it must be clearly labeled.
Healthcare and Medical
AI systems used in healthcare are classified as high-risk. Medical devices and diagnostic aids are subject to strict regulation.
The healthcare sector is particularly affected by the AI Act because medical devices incorporating AI are automatically classified as "High Risk". This classification combines the requirements of the Medical Device Regulation (MDR 2017/745) with those of the AI Act, requiring rigorous dual compliance.
AI systems intended for medical purposes must comply with both the AI Act and medical device regulations (MDR).
Example applications:
- Medical image analysis
- Diagnostic assistance
- Patient triage systems
- Monitoring systems and clinical alerts
Dual compliance required
Medical AI systems must satisfy MDR/IVDR AND AI Act requirements: enhanced technical documentation, clinical evaluation, post-market surveillance, declaration of conformity, and specific CE marking.
Finance and Banking
Financial services using AI for credit scoring, fraud detection or trading are subject to specific obligations. Credit scoring is classified as high-risk.
The financial sector faces dual regulation: the AI Act for artificial intelligence aspects, and sectoral regulations (MiFID II, PSD2, Basel III) for financial aspects. Credit scoring systems are explicitly listed in Annex III as high-risk systems.
Concerned systems:
- Credit scoring and risk assessment
- Fraud detection
- Algorithmic trading
- KYC and identity verification
- Risk assessment and insurance pricing (potentially high risk)
Multi-regulatory compliance
Credit scoring systems must comply with all high-risk obligations. For other systems, ensure algorithmic transparency and decision explainability, also in accordance with GDPR requirements on automated decisions.
Agencies and Consultants
Are you a digital agency, law firm or consultant? Offer AI Act compliance services to your clients with our white-label platform.
The AI Act represents a major business opportunity for consulting firms, IT service companies, and digital agencies. With over 6,000 affected companies in France and imminent regulatory deadlines, the demand for compliance support is exploding. Position yourself as an AI Act expert with your clients.
Agency advantages:
- Your branding on all documents
- Multi-client management
- Custom compliance badge
- New revenue stream
- Customized audit reports with your branding
Target client types
Software publishers, AI startups, HR services, healthcare institutions, financial institutions, e-commerce retailers... All sectors need AI Act expertise. Our white-label solution allows you to offer a professional service without technical development.
The AI Act affects all sectors
Whatever your industry, if you use, develop, or market artificial intelligence systems in the European Union, the AI Act applies to you. Penalties can reach 35 million euros or 7% of annual turnover for the most serious violations.
Don't take risks: assess your AI systems' compliance now with our free audit tool. In just a few minutes, identify your systems' risk level and discover the actions to implement to be compliant before regulatory deadlines.
Not sure about your classification?
Our free audit determines your risk level and tells you exactly what you need to do to comply with the AI Act.
Our intelligent questionnaire analyzes your situation and provides a detailed report with concrete actions to implement.
Start my free audit