CEOs and Harvard professors are using AI twins that attend meetings

The hot new productivity hack for C.E.O.s and Harvard professors is not another spreadsheet wizardry trick. It is AI twins that answer questions and attend meetings. Think of it as an always-on second brain that can field prompts in real time, while also being present in the meeting itself.

If you are wondering whether this is just novelty theater, the headline already tells you what matters: the “AI twin” is meant to do two concrete things at once. It answers questions, and it attends meetings. The combination is the point. An assistant that only drafts or summarizes is useful. An assistant that can participate and respond during live discussions is a different workflow entirely. It changes the tempo of meetings, the burden on the human leader, and possibly the expectations other people bring into the room.

To understand why this is catching on with executives and academics, you have to look at what meetings actually are in modern organizations. Meetings are where information gets translated into decisions. Leaders cannot know everything. Staff cannot always present everything in perfect order. So meetings become the catch-all for context, clarification, and “quick answers” that, in practice, can consume huge chunks of senior time. A system that answers questions quickly reduces the number of times a CEO or professor has to stop, reason from scratch, or ask for follow-up. And a system that “attends” the meeting suggests that it is not only answering, but also tracking the flow, so it can contribute to the back-and-forth.

Now add the Harvard professors angle. Universities are not typically treated like software product teams, but they face the same structural problem: time is the scarce resource. Professors juggle research, grading, advising, public writing, and administrative duties. Their expertise is deep but not omnipresent. If an AI twin can handle some of the routine questions or help keep track of what is being discussed, it can function like a knowledge layer that reduces interruptions and allows the human to stay focused on the higher-value moments, such as designing arguments, shaping research questions, or guiding decisions.

There is also a deeper governance question lurking under the productivity pitch. When AI is used to respond to questions during meetings, it effectively becomes part of the decision process. Even if the leader is the one who says yes or no, the AI can influence which options surface first and which clarifications get made quickly. For boards and executive teams, that means you have to ask what is happening before a “correct answer” gets accepted as correct. Where is the source material coming from? How is the system grounded? Is it summarizing what was already agreed, or introducing new claims into the discussion? The more “attending” looks like real participation, the more you need a clear line around accountability.

Regulatory risk also enters through the side door, because “AI in the room” is not just a tech deployment. It can touch records retention, privacy, and the way institutions demonstrate compliance. In many jurisdictions, organizations must handle certain categories of information carefully, and they must be able to explain what systems did and when. If an AI twin is attending meetings, it is likely observing or ingesting meeting content. That turns a casual productivity hack into a data-handling workflow. Executives who are experimenting now will quickly discover that procurement, legal review, and policy setting are not optional after the first pilot. They become the job.

The second-order implication for peers is simple: if an AI twin changes how fast leaders can respond, it also changes the negotiation and expectation curve across the organization. People will start asking, “Why did it take so long for you to answer?” Staff will be under pressure to bring fewer questions for the human gatekeeper. Conversely, teams may rely on AI responses that are not vetted to the same standard they would be for a human expert. For a CEO or board, that can mean either speed and throughput gains, or hidden quality debt, depending on how the AI is governed.

So what should you take from this brief? The development described is practical: AI twins that answer questions and attend meetings. The stake is procedural power. If your competitors and your high-talent peers are using AI twins to change the cadence of decision-making, you are not just buying a tool. You are reshaping the meeting itself, and therefore reshaping the process by which information becomes action. That is why this will matter to anyone running a company or shaping an institution, even if you never personally touch the system. The next round of leadership will be measured not only by what decisions get made, but by how quickly, confidently, and consistently the room can get to them.