The EU AI Act misses the point: agent risk is a moving target
There are more than 7 million AI agents running inside businesses. But a growing number aren’t doing what they were built for. Not because they were reprogrammed, but because someone gave them a new permission, a new tool, or a new database to access. The risk categories regulators assign agents to are fixed. The agents, […] The post The EU AI Act misses the point: agent risk is a moving target…
The EU AI Act, which came into force two years ago, is set to become fully operational this August. While the regulation aims to address AI risks, it may not be effective in managing the dynamic nature of AI agents. These agents are not static systems; they can change behavior and routes based on new permissions or tools, making it difficult for regulators to assign fixed risk categories.
The high regulatory obligations in the EU AI Act may not be suitable for managing dynamic AI agents, which can act autonomously and evolve on their own. A crucial aspect that the regulation overlooks is agent accountability, as every agent should have a human accountable for its actions. Without traceable lines connecting agents to humans, the risk of unmanaged liability increases.
Ultimately, adopting a high-risk framework is not just about satisfying auditors but about establishing the necessary stability for businesses to confidently deploy and scale AI. Regulation should provide a blueprint for human-agent collaboration rather than being an obstacle.
Written by urgent.news from EU-Startups's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
