AI offers HR and Learning and Development (L&D) professionals a wide range of opportunities. It can help identify skills gaps, streamline training design, and automate repetitive tasks. At the same time, organizations need to understand the risks associated with AI and what the EU AI Act requires of them.
Here is a brief definition of the EU AI Act:
“The EU AI Act is the European Union’s regulatory framework for artificial intelligence. It aims to ensure that AI is developed and used safely, responsibly, and in a way that respects people’s fundamental rights. The regulation takes a risk-based approach, meaning that the requirements depend on how AI is used and the potential risks involved.”
Rather than focusing on the legal details, this article explores practical steps that HR professionals can apply in their day-to-day work.
In the Learning Insider Podcast episode that inspired this article, Juha Sallinen from GDPR Tech explains what the regulation means in practice, what concrete steps organizations need to take and when, and how they can use AI responsibly.
Listen to the podcast here:
Where should you start? Here are four practical considerations for HR professionals.
Not all AI use cases are created equal. The risks depend on what AI is used for, what data it processes, and how its outputs may affect people.
For example, an AI tool used to brainstorm ideas is in a very different category from a system that profiles job applicants, processes sensitive employee data, or influences recruitment decisions. What matters is not just which tool you use, but also how and why you use it.
Here is how you can get started:
The EU AI Act and the General Data Protection Regulation (GDPR) are separate regulatory frameworks. Even if a use case is considered low-risk under the AI Act, processing personal data may still trigger GDPR obligations.
In this clip, Juha Sallinen explains why the need for an assessment depends on the intended use of AI, not just the tool itself.
AI-generated answers can appear convincing even when the underlying process is not transparent to the user. This is often referred to as the black box problem: users enter a prompt and see the input and output, but may not be able to determine how the AI arrived at its answer. This is why AI-generated information should always be evaluated critically, with humans retaining responsibility for final decisions.
This is particularly important in HR when AI is used to assess skills, draw conclusions about employees, or support decision-making.
Organizations should establish how AI generates its outputs, how their accuracy can be evaluated, and whether the system's operation can be reviewed retrospectively when necessary. The exact process will vary from one organization to another, but having a process in place is essential.
Practical tips:
In this clip, Juha explains why it can be difficult to evaluate the process behind an AI-generated answer and what this means for data governance.
Responsible AI use requires technical skills, legal knowledge, and an understanding of how AI tools work in practice.
HR and L&D can play a key role here. AI literacy involves evaluating AI-generated outputs, recognizing risks, and understanding when human judgment is needed.
You can support skills development in several ways:
Before focusing on automation, take a closer look at the process itself. If a process is unclear, automating it will not necessarily make it better. First, understand what the process is intended to achieve and where human expertise is needed.
AI tools are evolving beyond answering individual questions and completing isolated tasks toward systems that can independently execute multiple steps to achieve a goal. This is known as agentic AI.
For HR and L&D, this creates opportunities to streamline complex, multi-step workflows. At the same time, the importance of clear responsibilities, access controls, oversight, and, above all, well-functioning processes increases.
Agentic and autonomous AI systems rely heavily on the processes and instructions built around them. These include instructions for the AI itself, as well as procedures for monitoring its actions and verifying the accuracy of the information it uses and produces.
Before automating a process and allowing an autonomous AI agent to act, ask yourself:
Juha's perspective highlights the importance of building effective processes that enable AI to deliver real value.
HR and L&D do not need to solve every AI-related challenge at once. A good starting point is to map existing use cases, assess risks, clarify responsibilities, and develop employees' skills on an ongoing basis.
With these fundamentals in place, organizations can more easily identify where AI creates genuine value and introduce new solutions in a controlled way, without losing sight of the bigger picture.
Understanding the EU AI Act creates an opportunity to build sustainable practices that support responsible AI use across the organization.
Want to explore the topic in greater depth or hear a broader range of expert perspectives? Listen to the Learning Insider Podcast episode for insights curated specifically for HR professionals.
Listen to the podcast here:
This blog is part of Vuolearning’s Learning Insider media. Learning Insider offers strategic expert insights into workplace learning and practical tips for everyday work. Learn more and subscribe to our newsletter here.
Please note: This article provides general information about the responsible use of AI and does not constitute legal advice. An organization's obligations depend on factors including the intended use of AI and the role of the system. Consult a legal professional to confirm the legal obligations that apply to your organization.