KEEP IN MIND

EU develops joint plan against AI-driven cyber threats

The action plan must gather member states and companies around new risks and defences. It does not mean that new requirements come into effect immediately.

Publicerad 18 July 2026, 18.39

AI generated illustration of a European meeting on digital security and AI.

The European Commission has presented an action plan for cyber security and advanced AI. The plan will bring together member states, companies and EU bodies around both new risks and the possibility of using AI to detect and stop attacks.

The commission describes how advanced models can be used to find vulnerabilities, automate attacks and increase the pace of cyber incidents. At the same time, the same technology can help defenders inspect code and detect suspicious activity.

The action plan is a political direction, not proof that a particular AI attack has occurred and not a new everyday requirement. Several parts are to be built later together with member countries and relevant businesses.

Därför spelar det roll

AI can make email scams, fake voices and targeted attacks more credible. For Swedish businesses, the practical lesson is not to trust a message just because the language, voice or sender looks correct.

Det här kan du göra

  1. Confirm urgent payments through a second channel.
  2. Use multi-factor login and keep permissions limited.
  3. Report suspicious emails and voicemails before someone takes action.