.png)
The AI Act is a European regulation designed to harmonise rules on artificial intelligence across all EU Member States. It is based on a risk-based approach, distinguishing between different levels of supervision depending on the use of AI. Some of the key points of this regulation are as follows:
- Prohibited AI: Certain AI applications will be outright banned, including those used for social scoring (inspired by the Chinese model) or behavioural manipulation.
- High-risk AI: These systems, used in sensitive sectors such as healthcare, justice, education and recruitment, will have to meet strict requirements in terms of transparency, governance and compliance with ethical rules.
- Transparency obligations: In certain cases, companies will be required to inform users when they interact with AI (e.g. chatbots, AI-generated images or text).
-min.png)
Companies must comply with numerous legal and ethical requirements when using AI. Key measures to anticipate include: AI governance: Implement internal policies and processes to regulate the use of AI tools.
- Personal data protection: Ensure compliance with the GDPR, in particular by avoiding the disclosure of sensitive information when using AI tools such as ChatGPT.
- Algorithmic bias control: Verify that AI systems do not produce discrimination (e.g., in recruitment, facial recognition).
- Training and awareness: Educate employees and users about the limitations and risks associated with AI.
Beyond regulation, AI raises fundamental questions in terms of ethics and democracy:
- Algorithmic bias and discrimination: Poorly trained AI systems can reinforce inequalities (e.g., racial bias in facial recognition).
- Anthropomorphism and confusion: Some users may attribute consciousness to AI, leading to risks of dependency or manipulation.
- Surveillance and individual freedoms: Without oversight, AI can become a tool for mass surveillance (e.g., China's social credit model).
The AI Act marks a crucial step in the regulation of artificial intelligence, but it is only the first framework. Companies must now take a proactive approach to anticipate regulatory and ethical developments. Between innovation and control, the future of AI depends on the ability of economic and political actors to establish clear, fair rules that respect fundamental rights.
-min.png)
