The Zoom CEO Wants to Send Your AI Agent to Meetings

Key insights

• Yuan splits work into two steps. Human-to-human and human-to-system. He wants AI to absorb the entire second step, not just deliver better meeting notes
• The sharpest pushback came from the host. If attending a meeting suddenly costs nothing, an organization may hold more of them, not fewer
• Yuan contradicts his own clickbait headline. He says AI-run meetings without humans are many years away and that face-to-face still matters
• The context is sharper than the interview suggests. The broadcast opened with a report that AI will reshape 50 to 55 percent of jobs within three years

In Brief

Eric Yuan founded Zoom in 2011 after leading development of WebEx at Cisco. In April 2026 he sat down in a Fox Business studio with host Liz Claman and laid out what he thinks comes next: you won't need to show up to meetings anymore. Your AI agent shows up for you, throws in the ideas you would have thrown in, and handles all the follow-ups afterward.

Claman wasn't sold. She asked two good questions. First: does AI make meetings more efficient, or does it just lead to more meetings, because suddenly attending costs nothing? Second: how much gets lost in translation when an AI handles follow-up instead of a human?

Yuan answered both, but the interview also revealed something the clickbait headline hides: Yuan himself says AI-run meetings without humans are many years away, and that direct face-to-face still matters.

The core claim: meetings without the attendees

Picture a typical work meeting today. Five people dial in, talk for an hour, and each spends another half hour on cleanup: notes, emails, CRM updates, status pings in Slack.

Yuan's pitch is that AI agents can handle both the meeting and the cleanup for you, as he puts it in the interview:

"There'll be a point where people don't have to attend the meetings. An agent can attend them in place and communicate during the meeting, throw in ideas that I myself might throw in, if the agent knows me well enough."

That's a strong claim. You're not just skipping the follow-up. You're skipping the meeting itself. An AI that knows you well enough stands in for you in the room.

The two-step model behind it

Yuan lays out the logic like this:

Today we do step two manually, and that's where most of the friction lives. Yuan wants AI to absorb the entire second step, so it becomes, in his own word, "frictionless."

His own example: you finish a Zoom call with a prospect. Before: you log into the system, record the meeting, and handle follow-ups yourself. After: the AI does all of that the moment the call ends.

This isn't "AI writes better meeting notes." This is AI replacing the entire aftermath.

The host's two objections

Claman, who has hosted The Claman Countdown for years, heard something that didn't add up. She asked two questions worth looking at one at a time.

Objection 1: drop the cost, get more meetings

Claman's first pushback:

"Does A.I. make us more efficient during meetings or possible it risks making meeting more frequent because the cost of attending drops?"

This is a classic Jevons paradox objection. The paradox is named after economist William Stanley Jevons, who noticed in 1865 that as steam engines became more efficient, England burned more coal, not less. When something gets cheaper to use, we often use more of it, not less.

If attending a meeting suddenly costs you neither time nor attention, what stops an organization from scheduling twice as many?

Yuan's answer is that you can send your AI agent and let it follow up for you. But that doesn't actually answer Claman's question. It answers a question about you specifically. It doesn't answer the question about the system around you.

Objection 2: things get lost in translation

Claman's second pushback:

"I still feel there would be a disconnect. Sure we get summaries of meetings from A.I. That is great. Helpful. But there are always follow-up questions, and some things are lost in translation."

Claman points out that summaries are nice, but there are always follow-up questions, and some things slip through.

Yuan's response is that manual follow-up isn't always accurate either, and that AI can actually drive more precise and measurable follow-ups.

That's a fair point, but it assumes the AI can catch what was said between the lines. That's what Claman isn't convinced about.

What Yuan himself says about the timeline

Here's where it gets interesting. Claman asks directly: how close are we to AI running meetings without humans?

Yuan's answer was that it will take "many years," because "human-to-human direct like this is very important," and that he himself wants to "talk with you in person."

That's a milder version than the clickbait headline suggests. Yuan himself thinks it will take many years, and he openly says he wants to talk with people in person. The headline "AI agents could REPLACE meeting attendance" starts to stretch when the CEO himself says it will take many years and that the in-person meeting still matters.

The context the broadcast opened with

Right before the interview, Fox Business ran a report that set the tone: AI will reshape 50 to 55 percent of jobs within three years, and replace 10 to 15 percent within five. Yuan was asked whether software companies are in trouble, and answered that AI is instead creating new services, jobs, and opportunities.

That stance, that AI creates more than it destroys, is the standard CEO answer, and it's worth noting. But the number from the report is sharper than Yuan's calm reply. If more than half of all jobs get reshaped within three years, that isn't business-as-usual, no matter how relaxed you make it sound.

The real shift

Even if the clickbait headline stretches, there's something real in what Yuan describes. The shift isn't that AI replaces meetings. The shift is that the aftermath becomes invisible.

Today:

1. You have a meeting
2. You spend time afterward documenting, updating systems, writing emails, coordinating next steps
3. Step 2 is the bottleneck

What Yuan describes:

1. You have a meeting (possibly with an AI agent standing in for you)
2. AI handles everything afterward
3. The bottleneck moves to how well the AI can catch what was said and follow up on it

Whether that's good or bad depends on Claman's two questions: does the organization stay disciplined about meeting count when attendance is free? And is the AI good enough to not miss what was said between the lines?

Both questions are unanswered. Yuan himself says we have many years to figure it out.

Glossary

Sources and resources