The dawn of a new era: unpacking the EU's groundbreaking AI Act

Blog post
Arash Hajikhani,
Heikki Ailisto

In a historic move on 13 March 2024, the European Parliament cast a decisive vote, approving the Artificial Intelligence Act (AI Act) with an overwhelming majority of 523 votes in favour, 46 against, and 49 abstentions. This pivotal moment marks the first instance of comprehensive, binding rules for AI being set on a global scale. With Europe at the helm, the world watches as a new standard for the regulation and implementation of AI technologies is established.

At the heart of the AI Act is a risk-based strategy, designed to mitigate the potential dangers AI systems might pose to society. The legislation identifies outright bans on AI systems that could lead to cognitive behavioural manipulation, especially targeting vulnerable groups, enforce social scoring, or facilitate biometric identification and categorisation without consent. Real-time remote biometric identification systems, such as facial recognition, are also prohibited, reflecting the EU's commitment to protecting individual freedoms and privacy. Furthermore, automatic recognition of emotions is prohibited.

High-risk AI systems will be under the microscope

The AI Act delineates a category of "high-risk" AI applications, encompassing areas as varied as critical infrastructure, education, employment, and law enforcement. These systems will undergo rigorous assessment prior to market release and will be continually monitored throughout their lifecycle. This ensures that technologies influencing essential aspects of life, from educational opportunities to judicial fairness, are used responsibly and ethically.

Importantly, the legislation empowers individuals, granting them the right to file complaints about AI systems. This ensures a level of accountability and provides a mechanism for redress, reflecting the EU's prioritisation of citizen welfare over technological advancement.

Generative AI technologies, such as ChatGPT, while not classified as high-risk, are required to adhere to specific transparency and copyright compliance standards. These measures, including the obligation to disclose AI-generated content and prevent the creation of illegal material, aim to foster an environment of trust and safety in the digital space.

Does AI Act stifle tech innovation?

With the anticipated formal adoption of the AI Act by May or June, a phased implementation will follow. This includes an initial six-month period for member states to prohibit banned AI systems, with further provisions rolling out over the following two years. Non-compliance carries significant penalties, with fines reaching up to 35 million euros or 7% of a company's global annual turnover, highlighting the seriousness with which the EU views these regulations.

Like with all significant legislation, some concerns and criticism have been voiced, particularly regarding the potential for the AI Act to stifle tech innovation. Representatives from the SME sector have expressed concerns about small and medium-sized companies becoming risk-averse, which could halt the uptake of AI technology until the practical consequences of the AI Act are known, e.g., through first court decisions. The EU is actively communicating about the AI Act and promoting research and the uptake of AI with various instruments and policies.

The benefits of generative AI need to be distributed equitably

In this context, the commentary made from a high-level leadership roundtable organised by EIT Digital adds valuable depth to our understanding of Europe's AI regulation landscape with VTT experts' presence. This session brought together experts from industry, academia, and policymaking to discuss the transformative impact of Generative AI in Europe and to identify appropriate action points.

There's an early indication of a brain drain in the EU, which signals the urgent need to attract and retain talent.

Key highlights from the contributions include VTT's perspective where we emphasised the importance of balancing regulation with innovation to avoid stifling progress, particularly for smaller startups. There's an early indication of a brain drain in the EU, which signals the urgent need to attract and retain talent. This requires a delicate balance, providing time for training and transition while ensuring that regulations do not unduly burden the very innovations that could drive Europe forward.

The discussions emphasised the need for investments that bolster the synergy between human capabilities and AI technologies, ensuring a smooth transition into the AI era by promoting complementarity rather than complete substitution of human jobs. This approach aims to maintain a balanced workforce while navigating the transformative potentials of AI.

Moreover, the necessity to distribute the benefits of generative AI equitably was stressed, especially as it automates higher-order creative and analytical functions. Developing frameworks to ensure ownership rights and fair compensation is crucial in protecting individual rights while fostering innovation.

The journey towards trusted AI has only just begun

The AI Act represents a monumental step forward in the governance of AI technologies, setting a global precedent for responsible innovation. As expressed by the European Parliament, the overwhelming support for the Act underscores a collective commitment to protecting citizens from the potential harms of AI, while also providing clear guidelines for businesses navigating this new regulatory landscape. Europe's approach — regulating "as little as possible — but as much as needed" — balances the promise of AI with the imperative to safeguard human rights and democratic values.

As we stand on the cusp of this new era, the EU's leadership in AI governance not only protects its citizens but also charts a course for the responsible development and deployment of AI technologies worldwide. The journey towards trusted AI has begun, and Europe leads the way, forging a path others are likely to follow.

Arash Hajikhani
Arash Hajikhani
Research Team Leader
Heikki Ailisto
Heikki Ailisto
Research Professor
Our vision beyond 2030

We commit ourselves to focus all our energy and expertise on those technological challenges where we can make the biggest possible impact.