A quantum internet needs repeaters that can extend entanglement without copying a fragile quantum state. In 2026, researchers demonstrated a metropolitan-scale repeater over deployed fibre and, in a separate experiment, sent entangled photons through live telecom fibre carrying ordinary data. Together, the results show what changed and how far a network still has to go.
Send a fragile quantum signal down an optical fibre and it fades, the way any light dims over distance. For the normal internet that is no problem: every so often an amplifier reads the signal and shouts a fresh copy down the line. For a quantum signal, that fix is forbidden. A core law of physics says you cannot copy an unknown quantum state, so you cannot amplify it. Past a few dozen kilometres, the message is simply gone.
That loss problem is one reason the quantum internet has remained largely experimental. A quantum repeater works around it by creating entanglement across shorter links, storing those links in quantum memories, and joining them. Practical multi-node chains still do not exist, but 2026 experiments moved important repeater functions onto metropolitan fibre.
The United States, Europe, and China are funding quantum-network programmes, while telecom operators and national laboratories are testing links. The important shift is not that a national network has arrived; it is that repeater hardware and ordinary telecom infrastructure are beginning to meet in realistic settings.
24.4 km
of live, internet-carrying city fibre that entangled photons crossed at over 94% fidelity, sharing the cable with normal traffic.
Here is what happened
A metropolitan repeater reached 14.5 kilometres. A USTC-led team distributed heralded entanglement between two solid-state quantum memories separated by 14.5 kilometres of deployed fibre. The Nature Photonics result used time-division multiplexing and certified Bell non-locality, a demanding test of the quality of the shared state.
Multiplexing attacked the rate problem. Instead of waiting for one entanglement attempt to complete before trying again, the system used many time slots, improving the chances of success while the memories held the state. That architecture is important because useful repeaters need both high-quality entanglement and workable distribution rates.
Entanglement also survived on live telecom cable. In a separate Northwestern experiment, entangled photons crossed 24.4 kilometres of commercial fibre between Evanston and Chicago while the same cable carried high-capacity internet traffic, preserving more than 94% fidelity. This was not a complete repeater; it showed that quantum and classical traffic can coexist on existing infrastructure.
Two complementary barriers moved at once. The repeater experiment advanced memory-based entanglement distribution, while the Chicago test advanced infrastructure compatibility. Neither produced an end-to-end quantum internet, but together they narrow two different engineering gaps.
How it works
Entanglement is the resource. Entangled particles produce correlations stronger than classical physics allows. Those correlations can support quantum key distribution, connect quantum processors, and link precision sensors, but they cannot be used to transmit usable information faster than light.
A repeater is a relay, not an amplifier. Because you cannot copy a quantum state, a repeater instead breaks the journey into short hops, entangles each hop separately, then stitches neighbours together in a step called entanglement swapping, using a quantum memory to hold each link until its neighbour is ready. String enough hops and entanglement reaches across a country.
The hard part is holding on. Quantum memories must preserve fragile states while neighbouring links are created. The 14.5-kilometre repeater result addresses that problem; the separate Chicago demonstration shows that entangled photons can share noisy, traffic-carrying fibre. A useful network will need both capabilities in the same scalable system.
Why it matters
New security primitives. Entanglement-based protocols can reveal attempts to intercept the quantum channel because measurement disturbs the state. That does not make an entire network invulnerable: endpoints, software, authentication, and implementation remain attack surfaces.
It could ride the fibre we already have. The biggest surprise of 2026 is that entanglement survived on live commercial cable. If a quantum network can share existing telecom fibre instead of demanding its own dedicated lines, the cost of building one drops enormously.
It networks the quantum computers themselves. Just as the internet made many computers more than the sum of their parts, repeaters could link separate quantum machines into one larger system, and connect ultra-precise quantum sensors and clocks across cities.
Whoever builds it sets the rules. A quantum internet is strategic infrastructure, for secure government and military communication, for finance, and for national security. That is why this is a race, not just a science project, and why the countries and companies that own the fibre and the know-how are moving now, before the standards are set.
The honest catch
These are first building blocks, not a network you can log in to.
Still a few links, not a chain. The demos connect two nodes, or beat a clock over one hop. A useful quantum internet needs many repeater nodes chained together, which no one has yet shown end to end.
Rates are still low. The entanglement distribution rates demonstrated today remain far below what many practical applications would require. Increasing useful rates while preserving fidelity and memory lifetime is a major open engineering problem.
Years of engineering. Quantum memories, single-photon detectors, and the timing to synchronise it all are finicky and often very cold. Turning lab demos into an always-on metro network is a long haul.
EDITOR'S TAKE
The quantum internet has remained a promise because scalable repeaters require several hard capabilities at once: memories, multiplexing, entanglement swapping, low loss, and reliable operation over deployed fibre. In 2026, a 14.5-kilometre metropolitan repeater certified high-quality entanglement, while a separate Chicago experiment showed quantum traffic sharing live telecom cable. The experiments solve different problems and should not be conflated. Watch for a multi-node chain that combines repeater memories with traffic-carrying infrastructure at rates high enough for a real application. That is the point at which a physics demonstration starts to resemble a network build-out.
Quick questions
What is a quantum internet actually for?
Not for browsing faster. A quantum internet would distribute quantum states or entanglement between distant points. That could support keys whose interception is detectable, connections between separate quantum processors, and networks of precision sensors and clocks. It would be a specialised layer alongside the ordinary internet, not a replacement for it.
Why can't you just amplify a quantum signal like a normal one?
Because of a rule called the no-cloning theorem: you cannot make a perfect copy of an unknown quantum state. The normal internet keeps signals alive by copying and re-broadcasting them at each amplifier, which is exactly what quantum forbids. A quantum repeater gets around it without copying, by creating entanglement over short hops and then "swapping" it to join the hops together, using quantum memories to hold each link. It is a relay built out of physics tricks, not a louder speaker.
Is this the same as the quantum computers in the news?
Related but different. Those stories are about quantum computing, machines that calculate in new ways. This is about quantum networking, moving quantum information between places. They connect: repeaters could one day link quantum computers into a network, the way the internet linked ordinary ones. But a quantum internet could be useful for secure communication and sensing long before large quantum computers arrive.
Sources
Quantum internet leaves the lab (Northwestern University): the 24.4 km live-fibre experiment and its distinction from a repeater.
A metropolitan-scale multiplexed quantum repeater with Bell non-locality (Nature Photonics): the peer-reviewed 14.5 km memory-based repeater result.
Entanglement distribution coexisting with conventional optical communications (Optica Quantum): the peer-reviewed study behind the 24.4 km Chicago live-fibre result.
Metropolitan-scale ion-photon entanglement via a quantum network node (Nature Communications): multiplexed node work behind faster city-scale links.
What is a quantum internet? (US Department of Energy): plain-language background on repeaters and the quantum-network vision.
Frontier Signal explains frontier technology in plain English. Company and study figures should be independently verified. This is general information, not investment advice.
