AI agents and the AI Act after the Digital Omnibus: roles, duties, deadlines
Which AI Act duties apply to AI agents, and from when: provider or deployer, high-risk classification, AI literacy and the new dates after the Omnibus.
Author
Robert Pelikan
Founder of oQueo, author of the HOW TO series and co-publisher of VANGUARD magazine. He writes the articles on governance, security and regulation of AI agents.
About the author
Robert Pelikan founded oQueo, an Austrian agency for AI consulting and implementation, software solutions, web and digital. He brings more than 30 years of experience in communication, marketing and branding, is the author of the HOW TO series, has been a podcaster since 2006 and is co-publisher of VANGUARD magazine.
On aegis-stack.com he writes the articles on governance, security and regulation of AI agents. Figures and legal statements in these articles are backed by sources. AEGIS is a product of oQueo that is meant to let organisations run AI agents in a controlled way on their own infrastructure.
Topics
- AI agents
- AI governance
- AI security
- EU AI Act
- NIS2 and the Austrian NISG 2026
- Marketing
- Branding
- Communication
Profiles online
Articles by Robert Pelikan
Least privilege for AI agents: permissions, accounts and credentials
How much access should an AI agent get? Its own account, rights scoped to the task, and credentials kept out of the prompt, short-lived and revocable.
Human in the loop: designing approvals that actually protect
Human approval of AI agent actions only protects when it is rare, specific and well designed. What AI Act Article 14 and automation bias research tell us.
NIS2 and AI agents: what the rules mean in practice
NIS2 never mentions AI agents, yet it covers them: access control, logging, 24-hour incident reporting and board duties, with Austria as a worked example.
Prompt injection: how attackers steer AI agents through content
Prompt injection hijacks AI agents through instructions hidden in emails and documents. Why filters fall short and which architecture limits the damage.
What is an AI agent? From chatbot to a system that acts
What sets an AI agent apart from a chatbot: a working definition, how it differs from assistants and automation, autonomy levels and documented limits.
When AI agents act on their own: documented incidents
Verified AI agent incidents from 2025 and 2026: data exfiltration, a deleted production database, poisoned supply chains. Sources, patterns, lessons.
Why an AI agent cannot check itself
Self-checks fail on shared context, correlated errors and approval fatigue. Why AI agent policy enforcement needs separate roles and signed approvals.
What is AI agent governance? Definition, risks and practice
AI agent governance decides what an agent may do, who approves it and how every action is logged. Definition, risks, EU law as of 2026 and first steps.