Data · AI · Law · Ethics

The magazine for decision-makers

AI Act: what really changes for companies that use artificial intelligence

The AI Act brings artificial intelligence governance inside companies. Here are the operational aspects to keep under control: risks, roles, skills and oversight.

by

·

·

1–2 minutes
L'emiciclo del Parlamento europeo a Bruxelles, dove si discute la regolazione dell'intelligenza artificiale

The European AI Act moves artificial intelligence out of the realm of pure innovation and into that of governance. For companies, the question is no longer just which tools to adopt, but also how to classify their use, document it and govern its risks.

Not all AI applications are equal

The core principle is a risk-based approach. An assistant that summarizes internal documents does not necessarily raise the same issues as a system involved in recruitment, credit or other high-impact processes. That is why the first step should not be filling in forms, but building an inventory of the AI use cases present in the organization.

Provider and deployer: understanding your role

A company may simply use a solution developed by a third party, or it may integrate, modify and distribute AI components in its own products and processes. The actual role shapes responsibilities and obligations. Buying a service therefore does not automatically transfer all governance to the vendor.

AI literacy: governance also depends on people

A corporate AI policy cannot be limited to technical security. Those who use these tools must understand their capabilities and limits: hallucinations, input quality, confidential information, the need for verification and the implications of automated decisions.

Five operational questions

  • Which AI systems do we actually use?
  • What data do they receive and where is it processed?
  • Which decisions do they influence?
  • Is there an owner for each use case?
  • Can we reconstruct how and why the system is used?

From compliance to design

The mistake would be to treat regulation as a final check, once the system is already in production. Security, transparency, human oversight, data quality and documentation work best when they are built into the design and procurement cycle.

In brief

For many companies the AI Act will above all be an exercise in organizational maturity: knowing which AI systems exist, who is responsible for them and what risks they entail. Good governance should not stop innovation. It should make it possible to innovate while knowing where the boundaries are.

This article is for informational purposes only and does not constitute legal advice.

Leave a Reply

Your email address will not be published. Required fields are marked *