The synchronization of brainwaves among students during class reflects how much they like the class and each other, a team of neuroscientists has found. The researchers followed a group of high school students and their teacher for an entire semester and recorded their brain activity during their regular biology classes using portable electroencephalogram (EEG) technology (pictured above). (Credit: Diane Quinn © 2015 Trevor Day School)
In a Nutshell
- Scientists laid out a roadmap for multi-brain neurofeedback, a system that reads two or more people’s brainwaves at once and feeds the sync back live to help them connect, learn, and work together.
- The approach sorts that feedback into three levels: split-second signal matching, shared mental pictures, and the overall health of a group.
- More sync is not always better, since it can signal conformity or one voice dominating, so the authors say the tool may work best by tracking the shifts between connection and disconnection.
Two people deep in conversation often slip into a hidden rhythm, their brainwaves rising and falling nearly in step. Scientists can now measure that rhythm as it happens and feed it back to both people in real time, and they believe a well-timed nudge could make therapy sessions, classrooms, and team projects run smoother. A clinician might one day spot the exact moment a patient’s brain starts drifting away from their own. A teacher might see when a struggling student has truly tuned in, or quietly checked out.
That future is still on the research horizon, but a new scientific framework published ahead of print in Trends in Cognitive Sciences lays out a roadmap to reach it. At its center is a technology called multi-brain neurofeedback, a system that reads brainwave patterns from two or more people at once and reflects that information back to them, through light, sound, touch, or even robotic movement, so the interaction itself can shift. Its authors argue the tool holds real promise but warn that the field needs to get far more specific about what, exactly, it is trying to train. The therapist and classroom scenarios above are the kind of uses they envision, not anything tested yet.
Underneath all of it sits a deceptively simple question: when two brains line up during a lesson or a therapy session, does that mean anything, and can steering people toward that alignment make their time together meaningfully better? According to the authors, the answer is a qualified yes, but only if scientists are careful about which layer of brain activity they target and why.

How Brain Sync Technology Reads Two Brains at Once
For years, neuroscientists could only watch what happened inside one brain at a time. More recent work records two or more people’s brain activity together, and a growing body of it links that alignment to social bonding, engagement, and learning. Researchers add an important caveat, though. Lining up does not guarantee an interaction is going well, and it can sometimes reflect people falling in behind a dominant voice rather than truly connecting.
Multi-brain neurofeedback pushes the idea further. Instead of just observing the sync after the fact, it measures the sync live and hands it back to participants so they can adjust on the spot. Several of the paper’s authors have spent the past decade building such systems, running them with thousands of pairs and groups in places as varied as museums and music festivals, sometimes alongside artists and musicians.
Three Levels of Brain Sync, Three Different Goals
A core contribution of the paper is a framework that sorts multi-brain neurofeedback into three levels, each with its own target, its own pace, and its own kind of feedback.
At the most basic level, the system tracks raw, moment-to-moment matching between brain rhythms, like two metronomes slowly settling into the same beat. Because this kind of sync happens in a flash, the feedback has to be nearly instant. In one experiment cited in the paper, learners picked up songs better after their brain rhythms were nudged into line with their instructors at a frequency tied to processing pitch. Most of that training happens below conscious awareness rather than as a skill someone practices on purpose.
A middle level digs deeper, watching whether two brains are building the same mental picture of a situation, with matching patterns of attention, prediction, and turn-taking. Research shows that when people team up to learn a rule together, their brains line up in ways that reflect those shared mental pictures. This sort of sync plays out over seconds rather than thousandths of a second.
At the highest level, the feedback aims at the health of the whole interaction. Does the group solve problems efficiently? Does it hold onto the cooperation and goodwill that keep people working together? One study found that how efficiently information flows through a group’s combined brain network can predict how well that team coordinates. Feedback here might sum up patterns over a minute or more, which could make it especially handy in therapy or team training.
Why More Brain Sync Is Not Always Better
One of the paper’s more surprising points is that more sync is not automatically better. Sync can even be harmful, the authors stress. A group where everyone thinks alike might just be conforming, with one person steering and the rest swallowing their own ideas. In therapy, the rhythm between two people is rarely smooth, and real breakthroughs often arrive only after tension, a rough patch, and repair. So the authors propose that the technology may work best when it targets the shifts between connection and disconnection, instead of chasing high sync as a goal in itself.
Real-world hurdles also stand in the way. Any system that gathers brain signals, crunches them, and sends back feedback takes a moment to do it. At the fastest level, where brain rhythms move quickest, even a tiny lag can snap the live link the system is trying to build.
Noise is another headache. When people move during a conversation, as they naturally do, that motion throws off electrical static that can masquerade as brain sync but is not. Scrub it out too aggressively, and the system risks erasing the lively, engaged moments it was meant to reward.
There is a sneakier catch, too. Simply knowing a brain is being monitored, and that a signal will report how in sync two people are, can push them to act more in sync on their own, separate from anything the system does. To tell real brain effects apart from this expectation, the authors call for strict comparison conditions, including fake feedback that looks just like the genuine kind.
Where Brain Sync Technology Could Go Next
Psychotherapy and education top the list of real-world uses the paper imagines. In therapy, a clinician could use the tool to catch subtle interpersonal patterns, the small moments where connection quietly frays, that might otherwise slip by unnoticed. In a classroom, it could flag when a student and teacher are truly on the same page versus when they have lost each other, allowing for better-timed help.
Workplaces are another target, where live readings across a group could one day guide team training in high-pressure jobs. Brainwaves need not be the only signal, the authors add. Heart rate, sweat, breathing, posture, gesture, speech rate, even word choice all shift together during a meaningful exchange. Brain readings can reveal what has not yet surfaced in behavior, the silent beat before someone speaks, but paired with those other clues they could paint a fuller picture.
Machines that read two brains at once already exist in the lab. What stands between today’s research and that future therapist or classroom is the harder job of figuring out what to do with everything those brains are saying, and proving the payoff is real.
Paper Notes
Limitations
This is a forum article that lays out a framework rather than reporting a single new experiment, so its job is to flag where the field still falls short. Live systems carry unavoidable processing delays that can break the precise timing biologically meaningful feedback depends on, especially at the fastest signal level. Cleaning the data in real time poses a bind of its own: filtering out movement-related interference can accidentally punish the active, engaged moments the system wants to encourage. Signal-level sync is also vulnerable to false positives, brain patterns that look shared but really come from both people taking in the same surroundings at the same time. Awareness of being inside a feedback setup can itself change how people behave, making it hard to separate the feedback’s true effect from expectation. The authors note that many foundational questions stay open, including which training targets fit which social settings, what timing limits effective feedback, and what ethical issues arise when nudging social coordination at the level of the brain.
Funding and Disclosures
According to the paper, Yafeng Pan received support from the Philosophy and Social Science Planning Project of Zhejiang Province, the National Natural Science Foundation of China, the Fundamental Research Funds for the Central Universities, and the Zhejiang Provincial Natural Science Foundation of China. Guillaume Dumas received support from the Institute for Data Valorization in Montreal and the Canada First Research Excellence Fund (IVADO; CF00137433), the Fonds de recherche du Québec (FRQ; 285289), the Natural Sciences and Engineering Research Council of Canada (NSERC; DGECR-2023-00089), and the Canadian Institute for Health Research (CIHR 192031; SCALE). Xiaojun Cheng was supported by the Humanities and Social Sciences Research Project from the Ministry of Education of China (No. 24YJC190006). Suzanne Dikker was supported by European Research Council grant ERC-CONS 101124369. The authors declare no competing interests. The paper notes that views expressed are those of the authors only and do not necessarily reflect those of the European Union or the European Research Council.
Publication Details
Authors: Yafeng Pan, Xiaojun Cheng, Guillaume Dumas, and Suzanne Dikker
Author Affiliations (as listed in paper): Pan is affiliated with the Department of Psychology and Behavioral Sciences and the State Key Lab of Brain-Machine Intelligence at Zhejiang University, Hangzhou, China, and the Zhejiang Key Laboratory of Neurocognitive Development and Mental Health. Cheng is affiliated with the School of Psychology at Shenzhen University, Shenzhen, China. Dumas is affiliated with CHU Sainte-Justine Research Centre, Department of Psychiatry, Université de Montréal, and Mila – Quebec AI Institute, Montréal, Quebec, Canada. Dikker is affiliated with the Department of Experimental Psychology at Ghent University, Ghent, Belgium, and the Department of Psychology at New York University.
Journal: Trends in Cognitive Sciences
Paper Title: “Multi-brain neurofeedback: what are we training for?”
DOI: 10.1016/j.tics.2026.05.007
Publication status: Published ahead of print, 2026; volume and issue numbers not yet assigned at time of publication.







