Niklas Rosén

When must a human step back in? The short answer: before a single mistake can spread. The last stretch before production is about what an agent remembers, where a person still needs to be in the loop, and how you actually know it's ready.

How autonomous should an AI agent really be? The short answer: less than the developer demo suggests.

What changes when an AI system stops answering and starts acting? The short answer: everything about how you need to think about risk.

With AI agents, we’re now moving quickly from pilots to production across business and sectors in society. And unlike task-specific automations, AI agents plan, decide, and act across multiple steps and systems on behalf of users and teams. This is not just a technical upgrade. It’s a new paradigm when it comes to security: We are no longer securing code - we need to secure behaviour.