One of the hardest parts of the EU AI Act is understanding how it applies in real life.
Most explanations focus on categories and definitions. But what companies actually want to know is:
"Does this apply to companies like mine?"
The answer depends on how AI is used — not just which tools are involved.
Below are practical examples of systems and companies, and how they are likely to be treated under the AI Act.
If you want to check your own situation directly:
Example 1: Hiring platforms and recruitment tools
Type of company: HR tech, recruitment SaaS, hiring platforms
Example systems:
- CV screening tools
- candidate ranking systems
- automated interview scoring
AI Act perspective: Often high risk
These systems directly affect access to jobs. Even if the AI is "supporting" decisions, it can influence outcomes in a meaningful way.
If your product touches hiring decisions, this is one of the clearest areas to take seriously.
Example 2: SaaS companies with AI features
Type of company: B2B SaaS, productivity tools, platforms
Example systems:
- AI summarization
- recommendation engines
- automated workflows
AI Act perspective: Depends on the feature
A summarization feature is very different from a system that ranks users or filters access.
The same company can have both low-risk and higher-risk features inside the same product.
If you're building software, this is worth understanding in more detail:
Example 3: Customer scoring and prioritization systems
Type of company: Fintech, marketplaces, platforms
Example systems:
- credit scoring
- customer prioritization
- risk scoring
AI Act perspective: Often high risk
These systems influence how people are treated, what access they get, and what decisions are made about them.
If your AI affects outcomes for customers, it moves into more sensitive territory.
Example 4: Internal AI tools inside companies
Type of company: Any company using AI internally
Example systems:
- internal copilots
- writing tools
- knowledge assistants
AI Act perspective: Usually low risk
These tools help employees work more efficiently but do not directly affect external individuals.
That said, companies should still be mindful of:
- data usage
- internal policies
- how outputs are used
Example 5: AI in customer support and chatbots
Type of company: SaaS, e-commerce, service companies
Example systems:
- AI chatbots
- automated support responses
- customer interaction tools
AI Act perspective: Typically limited risk, but can increase
If the system is purely informational, the risk is lower.
If it starts influencing decisions, resolving cases, or affecting outcomes, the situation changes.
Example 6: AI used in decision-making systems
Type of company: Various industries
Example systems:
- approval systems
- eligibility decisions
- automated workflows
AI Act perspective: Often high risk
The more a system replaces or strongly influences human decisions, the more important it becomes from a regulatory perspective.
What these examples have in common
Across all examples, one pattern is clear:
It's not about the tool. It's about the impact.
Two companies can use the same model and face completely different requirements depending on how they apply it.
If the system affects people, influences decisions, or changes outcomes — it deserves more attention.
How to map your own system
A simple way to evaluate your situation:
- Where is AI used in your business?
- Does it affect people or decisions?
- How strong is the impact?
- Is it internal or customer-facing?
These questions will usually give you a strong first signal.
If you want a structured way to do this:
Want to understand the full framework?
If you want a broader overview:
If you want to understand how risk levels are defined:
Next step: check your own use case
Examples are useful, but your situation is what matters.
The fastest way to get clarity is to map your actual use of AI and see where it lands.