Starting Thursday, August 1, the first European Union act dedicated to artificial intelligence will come into force. The rules are quite strict, and the fines for non-compliance are significant, reaching up to 7% of the turnover.
The Regulation (EU) 2024/1689 on Artificial Intelligence (Artificial Intelligence Act) comes into effect on August 1 and outlines the sectors and categories of businesses targeted, the types of AI systems affected based on the risk they present, the obligations for high-risk AI systems, and the penalties for non-compliance, according to the lawyers at Hațegan Attorneys.
The regulation will have a significant impact on sectors such as financial services (including banks), technology, pharmaceuticals, and medical services, utilities, and more. Non-compliance with the regulation could result in fines of up to 7% of the global turnover for prohibited uses and up to 3% for other violations. Regarding the protection of personal data, the AI Regulation complements the GDPR Regulation, not replacing or derogating from existing obligations but imposing new requirements to ensure this protection within new AI systems.
“The regulation adopted on June 13, 2024, will come into effect on August 1, 2024, and will be applied starting from August 2, 2026, except for a few chapters that will apply from 2025. Hațegan Attorneys recommends that companies review their AI systems to check if they comply with the regulations, map products and roles within the company, and assess risks and impact to determine the applicable requirements,” says Adoriana Azoiței-Frumosu, Senior Associate at Hațegan Attorneys.
The AI Regulation applies to the following categories:
- Providers who place AI systems on the market or put them into operation, whether they are established in the Union or a third country;
- AI system implementers with headquarters or presence in the Union;
- Providers and implementers from third countries, if the AI results are used in the Union;
- Importers and distributors of AI systems;
- Manufacturers of products that include an AI system under their brand;
- Authorized representatives of non-European providers;
- Affected individuals within the Union.
Classification of AI systems:
The act regulates four categories of AI systems based on the risk they present:
- Prohibited AI systems: Systems that recognize individuals and create social scores, deduce emotions at the workplace or in educational institutions, and other systems that affect the fundamental rights of EU citizens;
- High-risk AI systems: Systems used in critical infrastructure, employment, and public service provision;
- Limited-risk AI systems: Chatbots and systems that generate text, audio-visual content;
- Minimal or no-risk AI systems.
Obligations for high-risk AI systems:
These systems are subject to the most stringent requirements:
- Preparing technical documentation;
- Automatic logging of events and record-keeping;
- Providing clear usage instructions;
- Human oversight to ensure proper functioning;
- Implementing a quality management system;
- Preparing an EU Declaration of Conformity;
- Registering in the EU Database for high-risk AI systems.
In conclusion, the EU Regulation on Artificial Intelligence, an integral part of the Union’s digital package, aims to support innovation while simultaneously protecting health, safety, fundamental rights, democracy, the rule of law, and the environment. It will have direct application in Romania, making it essential for all involved entities to align their practices to avoid severe penalties and contribute to creating a safe and innovative digital environment within the European Union.
