Agentwashing: What AI Agents Really Mean for Work (and Learning) - Markus Bernhardt

As “AI agents” flood the market, most organisations are still in the same place: individualsexperimenting in private, inside isolated tools. That can boostpersonal productivity, but itrarely changes how work gets done across teams.

In this session, Markus Bernhardt reframes agents from tools to systems. Agents becometransformational only when workflows, communication, and decision structures evolve alongside them, so that agent-driven work becomes visible, auditable, and trusted. The session introduces a practical maturity ladder, moving from isolated use to shared, visible workflows to governance, and a simple way to separate real autonomy from agentwashing.
The focus stays practical and strategic, with an emphasis on where agents create value today, where they introduce risk, and what organisations need in place before they scale responsibly.

Key Topics Include:

• Why private agent use rarely changes how organisations execute work
• Agentwashing: how to tell autonomy from rebranded automation
• The maturity ladder: isolated use, shared and visible workflows, governed execution
• Decision rights: what agents decide, what humans decide, and what requires both
• Data contracts: the hidden prerequisite for agent-driven work