中澤りか：NTTはいかにAI時代の通信を再定義するか

Copper cables gave way to fiber. Circuit switching gave way to packet routing. Analog voice networks became broadband platforms that stream video to billions of devices at once. Each of those changes kept the same underlying design: light comes in through the fiber, gets turned into electricity to be routed and processed, and gets turned back into light for the next hop.

NTT’s IOWN initiative (Innovative Optical and Wireless Network) proposes to remove the electrical step. An all-photonics network would carry and process data as light from end to end. Rika Nakazawa, a senior executive at NTT who has helped shape the company’s innovation strategy, will present that work at the Tech for Impact Summit 2026 in Tokyo on April 26.

The Company Behind Japan’s Networks

NTT is one of the largest telecommunications companies in the world. Annual revenues approach $100 billion. Its research and development budget is comparable to those of the largest technology firms, and its history runs alongside the history of Japan’s digital infrastructure. The company operates the fiber networks, data centers, and communication platforms that more than 120 million people use daily. When a hospital connects a rural patient to a specialist in Tokyo, or a factory floor talks to its supply chain in real time, the traffic is usually on NTT equipment.

Networks that many millions of people depend on cannot be experimented with carelessly. They also cannot be left as they are. Data traffic from AI workloads, the number of connected devices, and the electricity consumed by digital infrastructure are all growing every quarter, and NTT has chosen to answer with a new architecture rather than incremental upgrades.

IOWN

In today’s networks, optical signals are converted to electrical signals at nearly every processing node. Each conversion uses energy, produces heat, and adds latency. Across billions of packets and thousands of nodes, those conversions are a large part of why data center electricity consumption is projected to rival that of entire nations within the decade.

IOWN replaces them with photonic processors that route, switch, and in some cases compute using light, with no electrical conversion. NTT’s published targets are 100 times the transmission capacity, one-200th the power consumption per bit, and end-to-end latency low enough for real-time interaction between people and machines thousands of kilometers apart.

NTT has demonstrated photonic components in laboratory and field trials. The IOWN Global Forum, a consortium that now includes Intel, Sony, and dozens of other technology companies, is writing the standards and specifications for commercial deployment. The roadmap puts initial commercial services in the mid-2020s and full-scale deployment by the end of the decade.

For executives weighing infrastructure investment, three things follow from those numbers. A network that uses far less energy fits corporate sustainability commitments and regulatory requirements. Near-zero latency supports applications such as remote surgery, autonomous vehicle coordination, and real-time digital twins of whole cities, which current infrastructure cannot reliably carry. And 100 times the capacity leaves room for the growth in AI-driven data generation that every forecaster expects.

Digital Twin Computing

IOWN is the physical layer. NTT’s Digital Twin Computing program is what the company intends to run on it: digital replicas of physical systems, including cities, ecosystems, human bodies, and economic networks, built, connected, and simulated in real time.

Manufacturers have used digital replicas of individual machines and production lines for years. NTT’s proposal is at a different scale. Using IOWN’s bandwidth and latency together with AI and sensing, the company describes digital twins covering whole urban environments, fed by data from transportation systems, energy grids, weather, population movement, and buildings, and updated continuously.

Disaster preparedness is the most obvious use in Japan. The country experiences roughly 1,500 earthquakes per year, has a typhoon season every year, and has been investing in resilience since the 2011 Tohoku earthquake and tsunami. A digital twin of a city’s infrastructure could simulate how a magnitude 7.5 earthquake would move through buildings, bridges, transport networks, and emergency response systems. Emergency services could rehearse events that have not happened yet. Urban planners could find weak points before they cost lives.

Climate adaptation works the same way. As sea levels rise and weather patterns change, cities need to know how their infrastructure will behave under conditions with no historical precedent, and a simulation running on infrastructure fast enough to process city-scale data in real time is one way to find out.

NTT Research, the company’s network of laboratories in physics, cryptography, and medical informatics, has been studying how digital representations of biological systems could speed up drug discovery, personalize treatment, and extend specialist expertise to places that lack it. Rural depopulation in Japan is producing communities where the nearest specialist is hours away. Telemedicine, real-time patient monitoring, and AI-assisted diagnostics on low-latency networks are the proposed answer.

From Carrier to Platform

For decades telecoms were utilities. They provided the pipes, other companies’ products flowed through them, and the carrier earned in proportion to volume. Margins compressed and there was little room to differentiate.

NTT is trying to become something else. Its investments in AI, quantum computing research, photonic processing, and digital twins are meant to combine into one capability spanning physical infrastructure, computation, and application services, with the network taking part in how data is generated, processed, and used rather than only carrying it.

That changes the position of companies building on top. AI developers, smart city vendors, autonomous systems makers, and precision medicine platforms have treated network infrastructure as a commodity input they did not control or think about. As their applications demand more in latency, bandwidth, and reliability, the infrastructure layer starts to decide what they can ship. Organizations that build relationships with next-generation infrastructure providers early will have architectural options that latecomers will find hard to copy.

Japan’s Digital Strategy

NTT’s work sits inside Japan’s national digital transformation effort, which has accelerated since the Digital Agency was established in 2021. Japan has an aging and shrinking population that needs productivity gains from automation, geographic exposure to natural disasters that needs resilient infrastructure, and a competitive position that increasingly rests on digital capability.

The government’s Society 5.0 program, a “super-smart society” integrating cyber and physical space, depends on the kind of network NTT is building. Smart cities, autonomous transport, remote healthcare, precision agriculture, and distributed energy management all need networks faster, more reliable, more energy-efficient, and more intelligent than today’s. NTT, as the country’s largest telecommunications provider and one of its largest R&D spenders, is carrying much of that load.

International executives trying to read Japan’s technology direction will find NTT’s partnerships, standards work, and infrastructure investments shaping the environment for every business operating in Japan, and increasingly across the Asia-Pacific region. The IOWN Global Forum’s membership list is a reasonable guide to who expects to be building on it.

What She Will Discuss at T4IS 2026

At the summit, Nakazawa will address how next-generation communications infrastructure opens paths to social impact at scale, and in some cases creates them. The conversation will move past specifications into strategy: what changes when the constraints that have held back telemedicine, disaster simulation, urban planning, and AI deployment are removed, what new kinds of collaboration become possible when latency and energy use stop being binding, and how leaders in other sectors should position themselves while the physical layer of the digital economy is rebuilt.

She joins a speaker roster that takes up these questions from several directions. Former Minister Taro Kono covers the policy side of Japan’s digital transformation. Cardano founder Charles Hoskinson brings the decentralized infrastructure view. GLOBIS founder Yoshito Hori delivers the keynote on entrepreneurial leadership. Kathy Matsui, general partner at MPower Partners, speaks on impact-driven venture capital. SmartNews CEO Ken Suzuki addresses AI and democratic discourse. Jesper Koll of Monex Group covers financial markets. Sota Watanabe of Startale and Ken Shibusawa of Commons Asset Management speak on Web3 infrastructure and multi-generational stewardship.

Every AI model, blockchain transaction, telemedicine session, and smart city sensor discussed at the summit runs on communications infrastructure. Nakazawa’s session is about what that infrastructure is turning into, and what the people who plan around it should expect over the next decade.

The Infrastructure Question

For the executives and institutional leaders at the summit, the state of communications infrastructure sets the limit on what their own plans can achieve.

Organizations investing in AI at scale are finding that model performance depends as much on data infrastructure as on the models themselves. Companies with sustainability commitments are finding that measuring and reducing their environmental impact takes sensor networks, real-time analytics, and computing that strain what they have. Governments promising digital inclusion are finding that connecting underserved communities requires networks that are cheap to operate as well as fast, which is where IOWN’s energy efficiency matters most.

The move from electronic to photonic infrastructure will take years. The forces behind it, energy costs, performance requirements, and data growth, are the same ones that drove every earlier network transition, from copper to fiber and from circuits to packets. NTT is betting that the result is infrastructure that can take on problems of the kind the summit exists to discuss, from disaster response to rural healthcare. Nakazawa, who works where NTT’s technical capability meets its strategy, will make that case in person.

---

The Tech for Impact Summit 2026 takes place on April 26 in Tokyo. Seats are limited and allocated by invitation. Request your invitation to join Rika Nakazawa and other global leaders shaping the future of technology, infrastructure, and impact.