Sokko & the EU AI Act

AI with people in control.

Give your team a clear way to use agents, with defined roles, careful access and a person responsible for decisions.

One job. A clear owner.Example workflow

Sokko Support

Drafts the reply

You review

Check facts, tone and policy

Ready to send

Your team makes the call

Make good habits part of the work.

Start with a specific task. Give the agent what it needs. Keep review and responsibility with the people who know the business.

Explore Sokko’s security

Give every agent a clear job

Name its role and write down its instructions. A support agent should know when to draft, when to act and when to ask a person.

Connect only what it needs

Choose the tools and credentials for the task. Sokko keeps agent workspaces separate and stores secrets in a vault.

Keep a person responsible

Build a review step into consequential work. Check outputs, correct mistakes and stop the agent when the task needs a human.

Your use case sets the responsibilities.

The AI Act takes a risk-based approach. Your obligations depend on your role and the system’s purpose. Using Sokko does not automatically make a deployment compliant.

European Commission: AI Act overview

Map the role and the risk

Using an AI system and placing one on the market carry different duties. Recruitment, education and access to essential services can fall under high-risk rules.

Help your team understand AI

Support AI literacy for the people using your systems. Cover their tasks, the limits of the models and how to spot mistakes.

Commission guidance on AI literacy

Make AI involvement clear

Check the disclosure rules for your chatbot or generated content. Requirements and exceptions differ for providers and deployers.

Commission guidance on transparency

The rules are already rolling out.

Updated 10 September 2026

Already applicable
AI literacy and the original prohibited practices have applied since February 2025. Article 50 transparency obligations apply from 2 August 2026.
High-risk rules: 2027 and 2028
Following the AI Omnibus, Annex III high-risk rules apply from 2 December 2027. Rules for high-risk systems in regulated products apply from 2 August 2028.

This is a practical overview, not a legal assessment of your use case. Review the official guidance and get advice for regulated or high-risk uses.

Read the official application timeline

What will your first agent do?

Create an agent, connect your tools, and give it a task. Free AI models are included to help you get started.