AI Literacy Is Becoming Mandatory for Businesses

Article 4 of the AI Act: What Businesses Need to Know Now

With the introduction of the EU AI Act, providers and operators of AI systems face new obligations. Article 4 of the regulation requires them to ensure that all employees and contracted personnel working with AI systems possess sufficient AI literacy. What does this mean in practice, and how can businesses meet these requirements?

What Is AI Literacy?

According to Article 4, AI literacy goes beyond technical knowledge — it also encompasses an understanding of the opportunities and risks associated with AI systems. Employees should be able to use these systems competently and identify potential harms. For instance, developers need to understand the algorithms behind AI systems, while end users must be able to interpret and critically evaluate their outputs.

Consider this example: A company using an AI-powered applicant tracking system must ensure that HR staff not only know how to operate the tool, but also understand how algorithms make decisions — in order to prevent potential discrimination.

Obligations for Businesses

To comply with Article 4, businesses are required to take proactive measures, including:

1. Training and professional development: Regular training programmes tailored to the specific needs of employees.

2. Ensuring technical understanding: Development teams must have a thorough understanding of the architecture and functionality of AI systems.

3. Regulatory compliance: Businesses must ensure that their processes and systems fully comply with the provisions of the EU AI Act.

Accounting for Individual Factors

Assessing AI literacy depends on a range of factors, including employees' experience and educational background as well as the context in which AI is deployed. For example, using AI in healthcare requires a very different skill set than in customer service.

Implications for Providers and Operators

These obligations apply regardless of whether a company develops AI systems in-house or merely deploys third-party solutions. Providers are responsible for development, while operators use externally developed systems under their own responsibility. Both roles demand a high degree of accountability and expertise.

Timeline

The EU AI Act was adopted by the European Council on 21 May 2024 and is expected to enter into force in June 2024. Following a 24-month transition period, businesses must be fully compliant by June 2026 at the latest.

Conclusion

Article 4 of the AI Act presents new challenges for businesses, but it also offers an opportunity to strengthen organisational AI capabilities. Starting training early and adapting internal processes is key to meeting the new requirements. Companies that engage with the regulation now will benefit in the long run through greater efficiency and increased trust in their use of AI.

Tip: Start assessing your current capabilities early and develop a training plan to meet the requirements of the EU AI Act.