the week the brain-chip race crossed a line.

This is issue one. I am starting this because the interface between the human brain and the machine is, to my mind, the most consequential technology almost nobody is watching closely, and it deserves a clear, weekly, no-hype read. Every issue tracks what is real, what is hype, and who is building it, anywhere in the world it is happening. One promise on sources: every claim here comes with a freely accessible link next to the in-depth report, so you can always read the story yourself.

The winner will not just have the best electrode. It will have the smartest path through the rulebook.

Everyone expects Neuralink to win the brain-chip race. They did not. China's regulator approved NEO, a coin-sized implant from Neuracle built with Tsinghua University, as the first invasive brain-computer interface cleared for commercial medical sale anywhere in the world. The twist is how they won. NEO sits on the brain's protective membrane rather than penetrating the cortex, a less aggressive design that cleared regulators faster. And China is not stopping at one device. Its Beinao No.1 system has now completed multiple implantations with thousands of hours of safe operation, and NeuroXess has broken ground on a Jiangxi super factory built for mass production. BCI now sits in China's latest five-year plan as a strategic future industry, next to quantum computing and humanoid robots.

The headline-grabber is China. The breakthrough is in a research paper. A study in Nature Medicine documented a man with ALS using a brain implant to speak and control his computer for over 3,800 hours across nearly two years, at home, with no researchers in the room, while holding down a full-time job. For years the question was can a brain implant work. This answered a harder one: can it work reliably, independently, for years. That is the difference between a lab demo and a product.

2026 is the year capital flooded in, and most of the biggest checks did not go to Neuralink. Science Corp, founded by a Neuralink co-founder, raised $230 million and now claims it could be first to bring an implant to market. In China, BrainCo raised about $286 million, the largest BCI round anywhere outside Neuralink, then filed for a Hong Kong IPO. Merge Labs, backed by OpenAI, came out of stealth with a $250 million seed to chase a non-invasive route. StairMed pulled in a round led by Alibaba with Tencent joining. The most telling entry is the smallest. Subsense is building a brain interface with no surgery at all, using engineered nanoparticles delivered by a nasal spray. It is early and unproven. It is also exactly the kind of name that did not exist on anyone's roster a year ago.

America still has the deepest bench, and this year it showed real clinical progress. Synchron closed a $200 million Series D and is on its FDA pivotal-trial path with a device threaded through blood vessels, no open-skull surgery, plus integration that lets patients control an iPhone or Vision Pro directly. Paradromics completed its first human implant of the 421-electrode Connexus system, aimed at restoring speech. Precision Neuroscience won FDA clearance for its Layer 7 cortical array. And Neuralink is moving toward high-volume production and an automated surgery, opening its first international site with the UAE-PRIME trial in Abu Dhabi. The honest caveat: no BCI is commercially for sale in the US yet. Every implant is still a clinical trial. Realistic timelines for limited availability run from 2028 to 2030.

Europe is not only the referee. It ships hardware, and its edge is in materials. Spain's INBRAIN ran the world's first human study of a graphene brain interface, used during tumor surgery to tell healthy tissue from cancer at micrometer precision, with zero adverse events. Germany's CorTec implanted its fully wireless Brain Interchange, and in one stroke patient it both delivered therapeutic stimulation and let him control a computer on the same hardware. Switzerland's Neurosoft is building an interface it says is a thousand times softer than the rigid arrays most implants still use. What Europe lacks is not science. It is late-stage capital.

Reading the brain is one race. Writing the rules is another, and here Europe leads. The EU AI Act already classifies BCIs as high-risk, and policymakers are now debating a dedicated neurotechnology strategy. But the US is no longer silent. Colorado became the first state to fold neural data into its privacy law, California followed, and Montana and Connecticut have joined, with Connecticut's protection taking effect on July 1, 2026. There is still no federal standard. The core worry is the same everywhere. Your brain never agreed to any terms and conditions, and neural data can reveal your health and mental state even when anonymized.