When Technology Reads Our Minds: The Urgent Fight to Protect Neurorights
Rafael Yuste is in his early 60s, with a striking resemblance to Pablo Picasso—if the iconic artist had worn wire-rimmed glasses and sported a neat, trimmed white goatee. Speaking in crisp, deliberate tones colored by his distinct Spanish accent, he walked me through a landmark experiment his Columbia University lab had run on mouse brains, focused specifically on the visual processing region of the cortex. His graduate advisor was Nobel Prize-winning Swedish neuroscientist Torsten Wiesel, whose pioneering work mapped how human visual systems interpret incoming sensory data. “He stumbled on a quirk by accident: the most powerful trigger for that visual cortex region is a pattern of high-contrast alternating dark and light bars,” Yuste explained, waving one hand back and forth in front of his face. “Imagine those fingers are strips of bright light set against total blackness. When I move them across your line of sight, that signal lights up nearly your entire visual cortex.”
The Groundbreaking Mouse Test That Rewrote Neuroscience Rules
The team first trained their test mice using those moving light bar patterns, projected onto a computer screen in front of the animals. When the bars slid up and down, that was a cue for the mice to sip water from a nearby tube. When the bars moved side to side, the mice learned they needed to stop drinking. The researchers used a precision laser system to scan through the mouse’s skull and track real-time brain activity, mapping exactly which individual neurons fired when the animal looked at each distinct moving bar pattern. “We could clearly identify the specific neurons that encode that exact visual stimulus,” Yuste noted.
Once the team cracked that neuronal code, they deployed a second holographic laser system to project pinpoint points of activation deep inside the mouse’s brain, triggering the exact same neurons that normally lit up for vertical or horizontal moving bars. “The make-or-break part of the experiment was when we turned off the physical screen entirely,” Yuste said. “It’s exactly like playing piano: you press different keys with specific fingers to make a melody. We were effectively ‘playing’ those pre-mapped visual patterns directly onto the mouse’s cortex. And when we did, the mouse behaved exactly the way we wanted it to.”
When the team activated the neurons tied to up-down moving bars, the mice licked the water spout as if the pattern was right in front of them. When they triggered the horizontal bar neuron set, the mice stopped licking immediately. The researchers had not only decoded the mouse’s visual processing in real time, but could fabricate a false perception the animal could not distinguish from reality. “The number of licks, the length of each lick, the exact delay before it started drinking—all of it was identical to the reaction when it saw the actual projected bars,” Yuste explained. “As far as the mouse can tell, those non-existent bars are physically there in front of its face.”
The demonstration laid bare the staggering power of the new neurotechnology: researchers could manipulate the mouse like a puppet, triggering any behavior they wanted by inserting custom perceptual signals directly into its brain. “If we can do this to a mouse today, there is no reason we cannot do it to a human tomorrow,” Yuste warned.
From Lab Breakthroughs to Mainstream Neurotech Applications
For two decades, researchers have used functional magnetic resonance imaging (fMRI) — which tracks oxygen-carrying iron in neuronal hemoglobin — to build increasingly precise maps of the mammalian cortex. Powered by massive leaps in machine learning AI, which can sift through vast datasets to draw statistical patterns and predictions, fMRI scans can now identify far more than basic visual signals: they can pick up on depressive thought patterns, nuanced emotions like envy and schadenfreude, and even reconstruct near-complete clips of movies a test subject watched, just by analyzing their brain activity. One 2020s study even scanned swing U.S. voters’ brains as they viewed photos and videos of presidential candidates, accurately predicting which candidates triggered anxiety or disgust, and which evoked positive feelings or empathy.
In just the last few years, neuroscientists have pushed past decoding images and emotions to decode full sounds, words, phrases, and even full flowing language. In a landmark 2023 demonstration, a woman named Ann Johnson who had been completely paralyzed for 18 years after a brain stem stroke regained the ability to speak: a 253-electrode grid implanted on the surface of her brain translated her neuronal signals into full sentences in real time, at a rate of 78 words per minute — roughly half the speed of natural casual conversation. The University of California team led by neurosurgeon Edward Chang paired the brain-computer interface (BCI) with an animated avatar of Johnson’s face, which spoke in her natural voice reconstructed from a 15-minute audio recording of a toast she gave at her wedding. The avatar’s facial expressions also shifted to match the subtle cues of her brain activity, translating her internal thoughts about smiling, frowning, or pursing her lips into visible emotional displays.
“They unlocked her,” Yuste said. “They cloned her language processing system directly into a computer — not her entire mind, but the speech and communication core. After we saw the results, Eddie called me up and told me he couldn’t sleep for days. He’d realized the full scope of both the incredible life-saving potential and the massive risks of this tech. It’s life-changing for people who are paralyzed, but stop to imagine what happens if you deploy this on a person for entirely different reasons. We hold a huge responsibility here. We’ve already built a machine that can decode every part of a person’s internal language. In 10 years, we will have a machine that can interfere with human thoughts the exact same way we altered the mouse’s perceptions in our lab.”
All brain-reading neurotech runs on the same core principle: it first records neuronal activity when a person completes a specific task (speaking, seeing, focusing, moving) to map what distinct electrical pulses, waves, or field signatures correspond to that behavior, then interprets those signals to read the user’s internal state. The more invasive the recording equipment, the higher resolution and more detailed the resulting data. Right now, fewer than 100 people on Earth have implanted BCIs like Ann Johnson’s, but a predictable trickle-down effect is rapidly bringing non-invasive alternatives out of labs and into public use. In summer 2023, a University of Texas team proved they could use regular fMRI to translate brain activity into full sentences: after test subjects listened to 16 hours of story-driven podcasts The Moth Radio Hour and Modern Love to train an AI model, the algorithm could accurately reconstruct the core gist of new audio clips the subjects heard, just by reading their brain scans. “When we got the system working, our first thought was, ‘Oh my God, this is kind of terrifying,’” lead computational neuroscientist Alexander Huth told Science at the time.
The Unregulated "Wild West" of Consumer Brain Scanning
Today, non-invasive wearable brain scanners are no longer confined to research labs: they are entering workplaces, and being sold to mass consumer markets for use in people’s homes. In a 2021 New Yorker interview, UC Berkeley cognitive neuroscience professor Jack Gallant — whose team works on building a complete functional map of the entire human brain — casually referenced a hypothetical future technology he called a “thinking hat.” He envisioned companies paying ordinary people $30,000 a year to wear the device, which would pair video-recording smart glasses with an array of neural sensors to capture every sight, sound, feeling, and experience the wearer encountered during their normal daily life. The scientific value of that volume of real-world brain data would be unprecedented for mapping full brain function, but the scenario instantly sparked a visceral dystopian reaction for many, including the author of this piece.
The trajectory is easy to imagine: cash-strapped graduate students drowning in student loans sign up as the first test subjects, then the tech spreads into the gig economy, where workers juggling multiple jobs jump at the chance to earn an extra $30,000 a year just for going about their regular routine, their brains being constantly scanned and mined for data. In effect, people would trade their raw neural activity to corporations and data brokers for cash, digital perks, or even basic website access — exactly the way most people now trade their personal browsing data and social media information for free online services. As Yuste points out, the human mind is the core sanctuary of personal identity. “You cannot treat that space like any other consumer dataset. You can’t just waltz in and start banking, buying, and selling brain data like it’s grocery store inventory.”
This is not far-fetched science fiction: early iterations of the “thinking hat” already exist on the market. Most current consumer devices sit on the scalp, and use electroencephalography (EEG) to pick up the tiny electrical fields generated by millions of firing cortical neurons, as they pass through the human skull. In recent years, EEG has become a widely used tool for tracking and even altering mental states, from focus and calm to stress and drowsiness. Neurotech firm Emotiv, for example, has been testing a combined EEG headset and earbud system for office workers, which monitors their brain activity throughout the workday to track attention levels, flag when focus drops or workers get distracted, and measure cognitive stress. The system is marketed as a productivity booster, telling workers to take breaks when stress spikes and attention falls. While the company initially promised data would be anonymized and only accessible to workers themselves, not their employers, that arrangement is almost certainly not permanent. As Emotiv CEO Tan Le predicts, “within the next five years, workplace brain tracking will be quite ubiquitous.”
EEG headbands are already widely deployed for truck drivers and mine workers, to flag extreme fatigue and prevent deadly workplace accidents — a use case with clear public safety benefits. Far more alarming is the example of U.S. firm BrainCo, which distributed its “Focus” EEG headsets to elementary school students in China to monitor their concentration levels, sending all the collected data to company servers accessible only to teachers, not the children or their parents. The program was suspended after a Wall Street Journal investigation exposed the practice. In Dubai, local police have used a technology called iCognative for years, which uses EEG to scan suspect’s brains and detect tiny involuntary spikes in recognition of objects or information tied to a crime. In one high-profile murder case, police showed a suspect a photo of the undisclosed murder weapon, and his brain produced an unmistakable recognition spike. When presented with the data, he confessed immediately — his own mind had effectively turned him in.
On the consumer market, EEG devices are mostly marketed to the wellness sector: there are headsets like the iBand, which claims to trigger lucid dreaming, and the Flow device, a home-use tool that sends targeted electrical pulses to mood, sleep, and motivation-regulating brain regions to treat anxiety and depression. Another popular EEG product, the BrainBit headband, promises to help users “unlock new potential” by tracking brain activity that can be used for unguided meditation, boosting schoolchildren’s attention spans, analyzing employee brain states at work, or even filtering potential dating matches based on your instinctive brain reactions. The device even advertises a feature that automatically broadcasts your current mood to your social media feeds. It also positions itself as a neuromarketing tool, collecting “neural insights” to capture consumers’ subconscious reactions to products for market research.
Major tech giants are already racing to enter this space: Meta and Snap are developing neurotech interfaces that let users trigger actions in virtual spaces just by focusing their attention on a virtual button, while Apple has filed a patent to integrate EEG sensors into a future version of its ubiquitous AirPods, giving the world’s largest consumer tech company direct, constant access to its users’ brain activity. None of these current wearables can yet decode full internal thoughts at the level the U.S. university teams demonstrated, but the pace of development is accelerating so fast that this capability is clearly only years away. One firm called Kernel has already released a headset that pairs EEG sensors with infrared light (essentially low-power laser scanning of the brain) to collect neural data of near-fMRI quality. The sleek, compact device looks like a BMX helmet, and is the closest real-world equivalent to Gallant’s “thinking hat” to date, with a full consumer version expected to launch within a decade.
The Birth of Neurorights: Building Guardrails for Consciousness
Yuste’s decades-long work in neuroscience was originally driven by a deeply personal mission: he wanted to find a cure for what he calls the brain’s natural “malware.” Early in his career, he worked in a psychiatric hospital in Madrid, treating patients with severe brain disorders, many of them paranoid schizophrenics. Some patients were so dangerous Yuste had to conduct interviews with a security detail present, including one exceptionally intelligent patient who pieced together Yuste’s home address just from small clues in his accent, and threatened to travel to his house and kill his father. That experience left a permanent mark on him: he realized something broken in that patient’s mind was driving him to turn against himself and society, and that doctors were powerless to stop it because they did not yet understand how the brain worked.
When Yuste and his team successfully implanted false visual perceptions directly into the mice’s brains, he saw both a breakthrough and a warning bell. The same technology that lets scientists map exactly where and how hallucinations form in the cortex could one day help cure schizophrenic patients. “I thought, ‘Oh my God, we can actually help these people now. We can go into their brains, reprogram their cortex, maybe cure them for good.’ But at the exact same moment, I realized that exact same method that heals a schizophrenic could be used to alter the mind of a completely healthy person.”
In 2017, Yuste convened a closed three-day summit at Columbia, bringing together 25 of the world’s top leading-edge neuroscientists, alongside clinical neurosurgeons, bioethicists, lawyers, tech industry pioneers, and AI experts to hash out the ethical and social consequences of neurotechnology. The group unanimously concluded that the rise of neurotech represents a fundamental unaddressed human rights crisis. “The brain generates the mind, and the mind is the core of what makes each of us human. If this isn’t a human rights problem, nothing is,” Yuste noted.
At that summit, the group coined the term neurorights, and proposed five new universal core rights designed to protect human consciousness from misuse and exploitation by neurotechnology:
The right to mental privacy: Brain data cannot be decoded, stored, shared, or sold without a person’s full, explicit, informed consent.
The right to identity: Clear legal boundaries must prevent neurotechnology from eroding a person’s inherent sense of self, ensuring no line is blurred between a user’s unique natural consciousness and external signals fed into their brain via interfaces.
The right to agency: Every person retains full control over their own decision-making, with no hidden, unconsented interference from neurotechnologies.
Fair access to mental augmentation: If brain-enhancing technologies become widespread, access must be open to all people, to prevent a permanent class schism that splits humanity between those who can afford premium neurotech tools and those who cannot.
Protection from algorithmic bias: All neurotech machine learning systems must be audited to eliminate the inherent social and cultural biases of their human creators and training datasets.
The group also drafted an ethical framework modeled after medicine’s Hippocratic Oath, dubbed the Technocratic Oath, for all scientists, entrepreneurs, corporations, and investors working in the neurotechnology sector.
Shortly after the summit, Yuste founded the non-profit Neurorights Foundation, an advocacy group that works directly with governments, policymakers, corporations, and the United Nations to codify global legal protections against neurotechnology abuses. The foundation has already secured major policy wins: after years of consultations with Yuste and his team, Chile became the first country in the world to enshrine neurorights in its national constitution in 2021, adding explicit amendments protecting brain data and brain activity. The move is unsurprising for a nation still grappling with the collective trauma of General Pinochet’s oppressive regime, where thousands of people were imprisoned, murdered, or disappeared for holding dissident thoughts and beliefs. The foundation now runs similar policy consultations with governments across Brazil, Uruguay, Mexico, Argentina, Spain, and the United States.
In April 2024, the state of Colorado partnered with the Neurorights Foundation to pass the world’s first standalone comprehensive neurodata law, extending existing state privacy protections to cover all brain data collected by non-medical consumer neurotech devices, noting that these systems “can collect and process information about an individual that the individual did not even know existed.” California followed shortly after by adding identical neuroprotections to its own state privacy laws. At the global level, UNESCO’s International Bioethics Committee has formally recommended reviewing and updating existing universal human rights frameworks, or even creating new dedicated human rights, to address the unique challenges posed by neurotechnology.
For Yuste, these legal guardrails cannot come fast enough. Technology has reached the very core of human consciousness, and the unregulated “extraction” model that has defined the big data era is now being applied directly to people’s minds. “Protecting neural data is not a problem we can deal with tomorrow,” says Jared Genser, lead lawyer for the Neurorights Foundation. “It’s a problem we need to solve today.”
The first independent study of consumer neurotech company privacy policies, published by the Neurorights Foundation the same month Colorado passed its neurodata law, analyzed 30 leading firms in the space. The results were alarming: all but one of the companies had unlimited unrestricted access to their users’ brain data, two thirds said they could share that data with third parties, and two already admitted to selling user brain data to external buyers. “Once you sell that data to a third party, that data is no longer bound by the original user agreement,” Yuste said. “Users can’t possibly have less protection than that. The companies are acting like this because there is zero regulation right now. They think, ‘Let me stake a claim to every bit of brain data I can get my hands on.’ It’s the Wild West, and every prospector is racing to claim land in the landscape of the human mind.”
Excerpted from The Vanishing Earth. Copyright © 2026 by James Crawford. Used with permission of the publisher, Bloomsbury. All rights reserved. If you buy products using links in our stories, we may earn a small commission to support our independent journalism.
