Neurotechnology in 2025

Key advances and breakthroughs across industry and academia

This is a companion article to “State of Neurosecurity 2025: Annual Report,” aiming to provide a quick overview of new neurotech developments in 2025.

Summary of Developments in 2025

2025 was a busy year for neurotechnology, with advances spanning invasive brain-computer interfaces, non-invasive brain monitoring, cutting-edge neural implants, and new therapies for neurological and psychiatric disorders. Researchers and engineers worldwide achieved milestones in brain-computer interfaces (BCIs) trials and introduced next-generation neural implants with unprecedented density and flexibility. Neuroprosthetics progressed toward restoring lost sensory and motor functions (including touch and vision), while neuromodulation technologies (closed-loop neural stimulator and focused ultrasound) opened new frontiers in treating diseases without invasive surgery. Neurotechnology for mental health made strides with clinical trials of deep brain stimulation for depression and improved non-invasive brain stimulation protocols. At the same time, many startups released non-invasive or minimally invasive neural monitoring tools (wearable EEG/fNIRS headsets, in-ear neural sensors), feeding into a growing ecosystem of cognitive enhancement and neurofeedback consumer products. Below, we delineate the major 2025 developments by key areas.

Brain-Computer Interfaces (BCIs)

Clinical milestones: In 2025 BCIs moved closer to real-world utility, especially for communication. Stanford scientists (Neural Prosthetics Translational Lab) demonstrated a BCI that decodes a person’s “inner speech” (the imagined words in a paralyzed patient’s mind) using implanted microelectrode arrays and machine learning. The researchers also addressed privacy concerns by training the decoder to ignore unintended thoughts by requiring a mental password b efore decoding, preventing accidental reading of internal monologue. Another breakthrough came from the BrainGate2 trial: an ALS patient used a BCI at home for over two years to communicate and work full-time. The implant uses 256 neurons for decoding, achieving up to 99% accuracy in translating his intended words (about 56 words per minute) and controlling computers. This demonstrated possibility of long-term reliability and usage over thousands of hours.

Industry milestones: Tech companies also embraced BCIs in 2025, indicating mainstream momentum. Notably, Apple introduced a new Brain-Computer Interface Human Interface Device (BCI HID) protocol in iOS, allowing implanted BCIs to serve as input devices for iPhones, iPads, and the Vision Pro AR headset. In August 2025, a public demonstration showed a man with ALS controlling an iPad entirely by thought using the Apple’s protocol and Synchron’s BCI. The Synchron Stentrode, inserted via blood vessels in the brain, captures motor intent signals and transmits them wirelessly to a computer. This was the world’s first instance of native thought-driven control, and Apple plans to roll out broader BCI compatability across its product ecosystems. Elon Musk’s Neuralink continued to expand ($650M Series E) and received Breakthrough Device Designation for speech impairment. Paradromic, another BCI company, performed its first human implantation this year and received an FDA approval for a human trial.

Next-Generation BCI Hardware: 2025 saw progress in the engineering of BCIs, making implants softer, thinner, and higher-bandwidth. Cambridge-based startup Axoft completed a first-in-human trial of its “ultrasoft” Fleuron electrodes, which are 10,000× softer than typical implant materials. These flexible electrodes were implanted in four patients during tumor surgeries and stably recorded high-resolution neural signals (single-neuron) from both the cortex surface and 1 cm deep into subcortical tissue. The soft electrodes had enough compliance to move with the brain’s pulsations, reducing signal drifting and tissue damage compared to rigid electrodes. This trial shows a possibility of more stable and biocompatible long-term BCI interfaces. Similarly, Spain’s InBrain Neuroelectronics reported positive interim results with its graphene-based electrodes. In a clinical study in Manchester, ultrathin graphene grids were placed on patients’ brains during tumor resections. The graphene electrodes’ flexibility allowed them to conform to brain tissue for clear signals even on complex cortical surfaces. This is the first human use of graphene neural interfaces, suggesting such materials can deliver brain recordings. Another headline breakthrough was the debut of BISC (Biological Interface System to Cortex), a fully integrated, paper-thin BCI chip developed by a DARPA-funded American academic consortium. Unveiled in Nature Electronics (Dec 2025), BISC is a 50-micron thin and flexible CMOS chip that sits like “a piece of wet tissue paper” on the brain. BISC packs 65,536 microelectrodes with 1024 simultaneous recording channels, powered and read out wirelessly at up to 100 Mbps. Initial surgical tests showed BISC can be implanted through a small skull opening and remain stable while capturing detailed cortical activity. Taken together, the advances in BCI software and hardware in 2025—from long-term patient success stories to innovations in materials and chips—suggest that practical neural interfaces are approaching feasibility.

Neuroprosthetics and Neuromodulation

Restoring Movement and Sensation: Neuroprosthetics (devices that replace or enhance lost neurological function) made significant headway in 2025. A highlight was progress in restoring vision via implants. Researchers demonstrated that directly stimulating the brain’s visual cortex (i.e. bypassing the eye, retina, optic nerve) can produce visual perceptions in humans with no sight. In a Science Advances paper, a team showed that blind volunteers consistently perceived phosphenes (flashes of light) and simple shapes corresponding to specific electrode activation patterns in the visual cortex. This points to a future where a digital feed (from camera or generated) could be encoded as electrical stimulation in the brain, restoring a form of vision for those with retinal degeneration or optic nerve injury. In a different approach, U.S.-based Science Corporation published 12-month PRIMAvera pivotal trial results showing a fully wireless subretinal photovoltaic retinal implant paired with camera glasses can restore functional central vision in patients with geographic atrophy from dry AMD.

For tactile sensation, researchers reported the longest human tests yet of intracortical microstimulation (ICMS) for touch feedback. A University of Pittsburgh group published 10-year data from five spinal-chord injury patients with microelectrode arrays in their somatosensory cortex. They delivered millions of tiny electrical pulses to these brain areas over the years and found it remained safe and evoked stable, localized sensations of touch in the subjects’ hands. Even after a decade, about 60% of the electrodes were still functioning, indicating the durability of such implants. This long-term safety milestone demonstrates the possibility of sensory neuroprosthetics such as brain-controlled robotic limbs with sense of touch.

For motor functions, systems that reconnect the brain to paralyzed limbs took a leap forward. In Switzerland, the ONWARD Digital Bridge project continued to enable natural walking in a man with spinal cord injury (first reported in 2023), and by 2025 several more patients have received similar brain-spine BCI implants. These systems use an implant in the motor cortex to decode the intent to move and then trigger patterned stimulation of the spinal cord below the injury, causing the legs or arms to move in synchrony with the person’s thoughts. Such BCI-bridged motor function is still in trial stages, but it shows the potential of neurotechnology to bridge neural lesions and restore voluntary control after paralysis.

Even more accessible neuroprosthetics arrived for spinal cord injury rehabilitation. In 2025, the U.S. FDA approved a stimulation device (Onward Medical’s ARC-EX) for at-home use in patients. The Onward ARC-EX is a noninvasive “stimulator pad” placed on the skin over the spine (transcutaneous electrical spinal cord stimulation). When combined with intensive rehab training, it delivers small electrical pulses that significantly improve arm and hand function in people with cervical spinal cord injuries. After initial clearance for clinic use in 2024, the FDA’s 2025 decision allows patients to continue stimulation exercises at home. This regulatory milestone indicates neurotech moving beyond labs and hospitals into everyday life.

Advances in Neuromodulation Therapies: Neuromodulation (interventions that modulate neural circuit activity to treat disorders) saw several important developments. Medtronic’s closed-loop DBS system (BrainSense Adaptive DBS), which automatically adjusts its electrical pulses in response to brain signals, received both CE Mark in Europe and FDA approval in 2025. The device adjusts stimulation on the patient’s real-time neural activity, reducing side effects and manual reprogramming. By October, more than 1,000 Parkinson’s patients worldwide had been implanted with this system. And a pivotal trial published in JAMA Neurology confirmed its long-term safety and superior clinical effectiveness over conventional DBS. TIME Magazine honored Medtronic’s aDBS as one of the “Best Inventions of 2025,” hailing it as the first closed-loop neuromodulation device to dynamically personalize therapy in real time.

Memory and mood were also targets for new implants. Building on a 2023 breakthrough in personalized closed-loop stimulation for depression, 2025 marked the launch of a new DBS trial for depression. In March, Mount Sinai Hospital in New York performed the first U.S. implant of a DBS device for treatment-resistant depression as part of a 25-site trial of Abbott’s DBS system. The study, “TRANSCEND,” focuses on stimulating the subcallosal cingulate cortex (a key depression circuit identified by basic neuroscience research) in patients with treatment resistant depression. After years of promising results in open-label studies, this is the first double-blinded, controlled trial to rigorously test DBS in depression. This effort, along with ongoing trials of closed-loop DBS for chronic pain and obsessive-compulsive disorder, highlights a new wave of psychiatric treatments with neuromodulatory devices.

Neurotechnology in Mental Health

Mental health care is being reshaped by neurotechnology, and 2025 brought several notable developments in this arena. As discussed, deep brain stimulation (DBS) for depression entered a pivotal trial phase. If successful, this could lead to the first FDA-approved DBS treatment for a psychiatric condition, analogous to how DBS is an established therapy for Parkinson’s. In parallel, academic trials of closed-loop neurostimulation for depression continued: a team at UCSF, for instance, has been refining a personalized system that detects neural signatures of depressive states and delivers calibrated stimulation only when needed. While results have yet to be published, the goal is a “smart” antidepressant implant that maintains mood balance automatically by modulating brain circuits.

On the non-invasive side, transcranial magnetic stimulation (TMS) saw innovations to improve its efficacy and accessibility. Clinics worldwide began adopting accelerated TMS protocols (Stanford Accelerated Intelligent Neuromodulation Therapy or “SAINT”), where patients receive high-dose theta-burst stimulation sessions multiple times per day over just 5 days. Pilot studies have shown it can achieve rapid remission in ~80% of severe depression cases, a dramatic improvement over standard once-daily TMS. In 2025, more data emerged on the durability of these results, and efforts were underway to gain regulatory approval for accelerated TMS as a first-line treatment in refractory depression.

Non-Invasive & Minimally Invasive Neural Monitoring

Neural monitoring tools become increasingly sophisticated and user-friendly in 2025. Wearable EEG systems became higher-density and easier to use, thanks to new dry electrode designs and integration with other sensors. For example, one innovative approach placed EEG electrodes in earbuds: IDUN Technologies unveiled an improved iteration of their in-ear EEG earbuds (the “Guardian” device) that can record brainwaves from inside the ear canal with comfortable dry electrodes. This design doubles as a normal audio headphone while invisibly tracking brain activity for applications like sleep monitoring and cognitive load tracking. The convenience of form factors like earbuds or headbands addresses one of EEG’s longstanding issues of discomfort and practicality. While these are marketed for research, companies like Wearable Sensing and g.tec released dry EEG headsets with 16–32 channels that can be donned in under 5 minutes, bringing research-grade EEG out of clinics and into offices or homes. Another modality gaining momentum is functional near-infrared spectroscopy (fNIRS), which uses near-infrared light to measure brain activity. The field is still improving signal processing techniques and reproducibility, but we saw launches of multi-modal products (e.g. fNIRS with EEG) in 2025.

Finally, minimally invasive neural interfaces that breach the body but not the brain itself made news. We already noted Synchron’s Stentrode BCI (inserted via blood vessels) enabling thought-controlled iPads, where its less invasive approach (no craniotomy needed) is a key selling point. Following that model, researchers are exploring other endovascular or endoscopic routes to the brain. A recently published study from ETH Zurich demonstrates a microbot capable of navigating complex cerebral vasculature. This hints at future “neural dust” particles or nanobots that might monitor brain physiology from inside blood vessels with minimal invasion or deliver targeted treatments.

Cognitive Enhancement

In 2025, cognitive enhancements through neurotechnology gained steam, both in research and the consumer market. A number of startups and device makers rolled out neurofeedback and brain stimulation gadgets claiming to boost focus, memory, or mental wellness.

Notably, several companies converged on multi-sensor headsets that provide real-time brain feedback to users during training exercises. For example, the newest Muse S “Athena” headband combined EEG and fNIRS sensors – making it the first consumer wearable to measure both electrical and hemodynamic brain activity. Other contenders in the space currently included Mendi (forehead fNIRS), FocusCalm (EEG), Sens.ai (EEG), and Neurosity’s Crown (EEG). These devices vary in approach, but all share the promise of personalized cognitive enhancement. The market for such “mental fitness” wearables grew in 2025, bolstered by a general cultural interest in self-optimization and some early scientific validations. However, there is still skepticism in the scientific community about the magnitude of benefits these consumer devices provide.

Aside from neurofeedback, nootropic brain stimulation continued to be explored. While the initial hype around transcranial direct current stimulation (tDCS) for cognitive enhancement has tempered (with mixed study results), we continue to see small trials and startup products utilizing tDCS and related techniques. Another trend is biohackers using neuropeptides imported from China to enhance brain functions. For example, people self-administer these with injection or intranasal delivery to enhance sociability or cognition. The space is largely unregulated, creating significant public health and national security risks.