The rise of artificial intelligence (AI) agents is undeniable in the United States, amid a torrent of applications and deployments across workplaces. European regulators, however, are taking a more cautious approach.
Earlier this month, the EU Commission imposed a first-of-its-kind ban on AI agents in its meetings without offering a reason or hinting at a broader ban. A note on the European Commission’s “Online Meeting Etiquette” slide starkly said, “No AI Agents are allowed.” [The Commission later confirmed a formal ban.]
Technology behind AI agents, which are generally used to transcribe, take notes, or even record visuals and audio during a video conference, isn’t covered by any specific legislation though AI models that power agents will have to abide by the region’s binding EU AI Act. The Commission is expected to drill down on the role of AI agents in its upcoming study of algorithmic management, a term referring to employee oversight via automated systems.
Aside from some nebulous language in a posting last month — “AI agents are software applications designed to perceive and interact with the virtual environment,” the Commission said — the EU declined to provide more details on its ban, leaving AI vendors scratching their heads and wringing their hands.
Srikrishnan Ganesan, founder and CEO of Rocketlane, said the most likely rationale is the EU is taking a deliberate approach with AI agents internally to ensure security and privacy.
“What’s unclear to me is does this mean ‘no recording of calls’? Or just ‘no AI agent’? What if this is just a video recording tool with no ‘AI’?” Ganesan said in an email. “As an entrepreneur building a business, I see huge value in our team recording customer and prospect conversations. This helps us focus on the conversation while the agent provides the notes, summaries, and actions from the call.”
“I find it unfortunate that there is a blanket ban on AI agents,” he added.
The restriction underscores the EU’s general measured approach to fast-developing AI technologies, starting with the one gaining the most attention this year, in AI agents. While OpenAI’s ChatGPT and Microsoft Corp.’s Copilot perform single tasks on request, AI agents represent an autonomous leap that impacts privacy and data sharing, according to Chris Mattmann, chief data and AI officer at UCLA.
“AI agents like Fellow sit and listen and take notes for you. However, where are those notes going?” Mattmann said. “Same goes with Microsoft 365 and turning on meeting summaries with LLM/notes after. Those agents and agentic AI provide no guarantees on what the meeting summaries and notes will be used for.”
For now, there is little debate the EU Commission’s actions could potentially have a paralytic effect on enterprises planning heavy AI agent use on the Continent if the ban is expanded beyond EU meetings. Roughly one-third of enterprise software applications will include agentic AI by 2028, compared with less than 1% in 2024, according to Gartner.
“Ultimately, the future belongs to AI agents built on secure, enterprise-grade foundations — where trust, governance, and measurable outcomes matter more than hype,” says Richard Bassett, vice president of CX automation and AI at NICE International.
