March 13, 2024 at 12:51PM
The European Union lawmakers approved the Artificial Intelligence Act, shaping the world’s first comprehensive set of AI regulations. The Act employs a risk-based approach, classifying AI applications into low and high-risk categories, with stringent regulations for high-risk uses. It also includes provisions for generative AI models and aims to influence AI regulations globally. The Act is expected to become law by mid-2026, with fines for violations reaching up to 7% of a company’s global revenue.
The European Union lawmakers have given final approval to the 27-nation bloc’s artificial intelligence law, the Artificial Intelligence Act. The act is expected to come into effect later this year and aims to regulate the use of AI with a human-centric approach. It divides AI applications into low-risk and high-risk categories, with stringent requirements for high-risk uses such as in medical devices or critical infrastructure.
Notably, the law also addresses generative AI models like OpenAI’s ChatGPT, requiring them to disclose the data used to train the systems and follow EU copyright law. Additionally, it adds extra scrutiny for the most powerful AI models to mitigate potential risks and harmful biases.
The EU’s regulations on AI are expected to influence other countries and global groupings, as they have been at the forefront of ratcheting up scrutiny of emerging technologies. Provisions of the AI Act will start taking effect in stages, with enforcement being carried out by individual EU countries, along with an AI Office at the EU level. Violations of the AI Act could result in fines of up to 35 million euros or 7% of a company’s global revenue.
Further legislation related to AI is anticipated after summer elections, as mentioned by Italian lawmaker Brando Benifei.