1
2
3
4
5
6
7
8
9
10
11
12

Green light for the EU AI Act: What B2B companies need to know now.

20.03.2024

With the approval from the European Parliament, the Artificial Intelligence Act (AI-Act) has cleared its final bureaucratic hurdle, paving the way for the world's first legal framework to regulate Artificial Intelligence (AI). But what are the exact intentions behind this legislative step by the European Union (EU), and what implications does it have for B2B companies? Here are answers to the key questions:

Why is there an AI-Act?

The AI-Act aims primarily to promote transparency, ensure data accuracy, and prevent discrimination by imposing strict documentation requirements on operators of AI systems.

Simultaneously, the EU law seeks to establish a uniform legal framework, providing reliability and fostering innovation within the EU. This includes granting startups and SMEs access to EU resources such as powerful supercomputers to enhance training and testing capabilities for newly developed systems.

Junge Menschen mit Smartphones in der Hand stehen vor einer Justitia-Statue. Symbol für Gerechtigkeit und Rechtsstaatlichkeit. Passend zum farblichen Erscheinungsbild der wob trägt eine junge Frau eine gelbe Jacke. | © wob AG
AI act ©wob AG

When will the AI Act come into effect?

Following approval from the European Parliament, the decision of the European Council is now the final step, expected in April 2024. Companies will then have a basic 24-month period to adapt their processes to comply with the new law. However, some provisions, such as certain bans, will take effect after just six months, with regulations on AI models for general-purpose use following in 12 months.

"The AI Act will bring extensive administrative tasks to B2B companies."

Who does the AI legislation affect?

The requirements of the AI Act primarily target companies and governmental bodies that offer or use AI systems in their operations. This means numerous European B2B companies will be subject to the AI Act's regulations.

What are the risk categories?

The EU's AI legislation categorises systems into four different risk categories:

  • Systems with minimal risk: e.g., spam filters or AI-driven video games.
  • Systems with limited risk: includes systems involving human interaction, such as chatbots acting as digital assistants on websites.
  • Systems with high risk: encompasses systems significantly impacting health, safety, or fundamental rights, e.g., AI systems used in candidate selection during hiring processes.
  • Systems with unacceptable risk: refers to social scoring systems or covert technologies like remote biometric identification.

What does the AI Act prescribe?

Essentially, the AI Act determines the purposes and conditions under which AI can be used in the EU. The higher the risk assessment, the stricter and more detailed the regulations to be observed.

While systems with minimal risk have virtually no transparency requirements, those with high risk face more stringent obligations. Operators must provide extensive information, such as how the AI system interacts with hardware and software. For systems with unacceptable risk, even stricter conditions apply: these applications are mostly prohibited, with few exceptions.

What's next for B2B companies?

The AI Act will entail extensive administrative tasks for B2B companies: reviewing their own processes, documenting used systems (technical), providing relevant information, and more. Companies need to act proactively and demonstrate readiness to adapt, as the exact implementation in member states remains unclear.

Simultaneously, the new legislation provides orientation and security for developing new AI systems. Additionally, it facilitates faster advancement of innovative ideas as the EU allocates resources for testing and developing new systems.

Change: Why proactive action is increasingly crucial.

The rapid pace of AI technological development (evidenced notably by the text-to-video model Sora) is unlikely to slow down in the future. Consequently, the AI Act will be regularly updated to keep pace with technological advancements. Therefore, B2B companies should always stay informed and be prepared to adjust processes to evolving guidelines.

Our conclusion: The AI Act as a guide in B2B.

The EU AI Act represents a significant step in regulating AI technologies, presenting new challenges and opportunities for B2B companies. Despite overcoming all bureaucratic hurdles, the exact implementation will unfold in the coming months. Adaptability, transparency, and a proactive stance remain key factors to fully leverage the benefits of this legislation and drive innovation.

Looking for a creative business partner merging extensive expertise with the strengths of artificial intelligence? Simply get in touch with us directly!