産業AIの主権
The intelligence that runs the fab, the grid, and the line — and who owns it. One day, one room, the leaders of the twenty-five-trillion-dollar industrial and semiconductor economy.
Hosted by the Industrial AI Federation · 産業AI連合
It would be an honor to welcome you to Tokyo — to consider, among equals, what sovereignty over industrial intelligence will mean for the next generation.
The Industrial AI Federation Convening since 2021
Since 2021, IIAC has gathered the heads of nations, fabs, and faculties who shape how intelligence is built into the physical world. The fourth edition moves to the industrial capital of the world — Tokyo — for a single, deliberate day on the question beneath every other question: who governs the intelligence that runs the factory, the grid, and the supply chain?
Expect the room to look like the global industrial-AI establishment — the chip giants and their equipment and materials suppliers, the hyperscalers, the founders of the industrial-AI stack — gathered in the city where much of it is built: Honeywell, IBM, Nvidia, Qualcomm, MediaTek; Panasonic, Tokyo Electron, JSR, Renesas, Furuno, CTC; FPT and the policy leads of METI and parliament; and the venture capitalists who fund them all. Attendance is limited. The room is the point.
Sovereign AI has crossed from op-ed into budget line. Nations, blocs, and firms are racing to own the intelligence that will run their industries — and 2026 is the year the money lands.
From India's IndiaAI Mission and BharatGen, to the UAE's state-backed AI strategy, to France and the EU's push behind Mistral and sovereign cloud, governments are funding domestic models, compute, and data governance rather than renting their future from foreign providers.
METI has committed roughly ¥2.35 trillion ($16B) to Rapidus for 2nm mass production by 2027, in partnership with IBM — the host nation's bid to reclaim the fab as sovereign territory, alongside national AI programs across Japanese industry.
Global semiconductor revenue hits a historic peak in 2026, AI chips now driving roughly half of it — the physical layer beneath every sovereign AI ambition, from the US CHIPS Act to the EU Chips Act 2.0 to Rapidus.
The firms and nations that move first will own the intelligence layer of the next industrial era. Be in the room when the next move is made.
Industrial AI is the intelligence that runs a fab, a grid, a supply chain, a factory line. It lives in process recipes honed over decades, in the heads of retiring engineers, in the defect-class logs of every wafer ever inspected. Sovereignty over that intelligence — owning it, on your own terms — is the question of the decade.
Compute, models, and fab capacity as instruments of state power — and the alliances deciding who builds, who depends, who decides.
The values, labor, and craft baked into the systems that train the next systems. Left unattended, the defaults get set by someone else.
The crown jewels of the firm — proprietary process data, the line's accumulated knowledge, the foundry floor itself. Sovereignty is the difference between making what you design, and merely designing it.
Because the firms that own the intelligence layer of their factories will own the next industrial era. Those that lease it back from a vendor-held cloud will pay forever, at a price that only goes one way. Sovereignty is not a slogan — it is the difference between making what you design and merely designing it.
Industrial AI is the intelligence running the fab, the grid, the line — yield optimization, predictive uptime, first-time-fix, the captured know-how of senior engineers. Sovereignty means owning the models, the data, and the deployment: on-prem, trained on your process, controlled by you.
Neurosymbolic models that reason like your experts. A shared sovereign frontier — Project Tapestry — that nations and institutions build on, then own. Knowledge-first architectures like DanaOS and the Cognitive Ontology, deployed inside your firewall, learning your craft and keeping it yours.
Models that read wafer-inspection logs and process recipes to lift yield and catch defect classes humans miss — without leaking the recipe to a vendor.
Field engineers guided by agents trained on decades of service tickets — closing the gap when the senior technician retires and takes the know-how with them.
Sensor-driven forecasts that schedule maintenance before a $200M tool goes down — owned by the operator, not rented from a cloud API.
The unwritten rules of the line — encoded into a neurosymbolic model your firm owns outright, so the next generation inherits the craft instead of rediscovering it.
A century ago this city taught the world that precision is a form of prestige. In October the question is sharper: when the global giants of the $25 trillion industrial economy walk into the same room, who owns the intelligence they leave with?
Tokyo October 8, 2026
This year's theme is anchored by Project Tapestry — the AI Alliance's open consortium for co-training a shared frontier model, from which nations and institutions build sovereign systems they own and control.
Tapestry is the through-line of the program: the technical substrate beneath national capacity, the cultural values embedded in the weights, and the industrial data that shapes them. Few subjects are more urgent, or more global, today.
A single day, a single track. The program is curated by the Program Committee around the three sovereignties — anchored by Project Tapestry. The detailed sessions are still being shaped; what follows is the shape of the day.
A curated program. Below, our keynote speakers and the committee whose networks shape the rest of the day. Around them, expect the chairs and CTOs of the global industrial-AI establishment — from Honeywell, IBM, Nvidia, and Qualcomm to Panasonic, Tokyo Electron, JSR, and Renesas — alongside the METI officials, parliamentarians, and venture capitalists who decide where the capital flows.
CEO & Co-Founder, Aitomatic Chief Architect, Project Tapestry
Built and led industrial-AI teams across Panasonic's global factories long before "industrial AI" had a name. Now CEO of Aitomatic, shipping DanaOS — the operating system for industrial AI — and Cognitive Ontology, the knowledge layer that lets sovereign models actually understand the factory floor.
President & CTO · Honeywell
Runs technology across Honeywell's $36B industrial portfolio — aviation, energy, buildings — and Honeywell Forge, the platform putting AI in charge of the world's factories and supply chains.
Managing Executive Officer & CTO · Panasonic Holdings
Chief Technology Officer of Panasonic Holdings and head of the Digital & AI Technology Center — the corporate engine putting AI into batteries, appliances, factories, and supply chains across the group, and the executive responsible for how Panasonic develops and deploys AI responsibly at industrial scale.
Director of AI Open Innovation, IBM Program Chair
Co-founded and co-chairs the AI Alliance — the world's largest open AI consortium — and runs IBM's open-source AI ecosystem, the engine behind Project Tapestry's industry coalition.
Bank of Baroda Chair Professor, IIT Bombay Founder, BharatGen
Founded and directs BharatGen — India's government-backed sovereign foundation-model initiative, covering the country's 22 official languages and bringing national-scale sovereign AI into Project Tapestry.
Senior Director of AI Research, Qualcomm
Runs AI research at Qualcomm AI Research, advancing the on-device and edge intelligence that puts sovereign models directly onto the chips in the world's phones, vehicles, and industrial sensors.
Nvidia Co-founder, OhmniLabs
Founded OhmniLabs (service robotics) and Katango (social AI, acquired by Google), pioneering physical and social AI at scale. Now at Nvidia, whose accelerated-compute stack underwrites every frontier model on the planet.
Renesas Electronics
Senior leader at Renesas — the world's #1 automotive microcontroller supplier, whose chips sit at the edge of nearly every car on the road and form the substrate of sovereign mobility AI.
MediaTek
Leads advanced wireless and AI connectivity at MediaTek — the world's largest smartphone-chip maker, shipping the silicon that puts intelligence into roughly two billion devices a year.
Senior Specialist · Tokyo Electron
Built his career writing the control software that runs Tokyo Electron's wafer-processing equipment — and now brings AI agents and SemiKong, the open-source semiconductor LLM he helped propose, onto the fab floor.
General Manager, Materials Informatics · JSR
Twenty-five-year materials-science veteran driving JSR's Materials Informatics Initiative — applying machine learning, with the Institute of Statistical Mathematics, to discover the photoresists and polymers that pattern the world's most advanced chips.
Aitomatic (Japan)
Heads Aitomatic's Tokyo operations, bringing knowledge-first AI to the global industrial giants whose process knowledge — accumulated over decades — defines what industrial sovereignty means in practice.
COO, Aitomatic (US)
Chief operating officer at Aitomatic and a core architect of DanaOS and the Cognitive Ontology — the operating system and knowledge layer that anchor Project Tapestry's shared sovereign frontier for the industrial world.
Director, Digital & AI Technology Center · President, Panasonic R&D Company of America
Leads Panasonic's Digital & AI Technology Center and heads the company's R&D operations across North America. Built his career on the UniPhier integration platform and LUMIX image-recognition systems before founding Panasonic's AI Solution Strategy — the playbook that put machine learning into hundreds of billions of yen of group business.
Minister · Government
Cabinet-level official shaping national policy on compute, capacity, and industrial strategy.
Full speaker roster published upon confirmation. Request an invitation to be notified.
A curated program is only as strong as the committee that shapes it. This year's committee draws from the semiconductor, systems, and research institutions building sovereign industrial AI — anchored in Project Tapestry.
The committee curates the program. To suggest work it should consider, write to pc26@iiac.events.
The Federation has convened IIAC since the first edition at Stanford in 2021 — bringing government, industry, and academia into the same room on the future of industrial intelligence. The fourth edition is the first held outside the United States, and is convened with the AI Alliance around Project Tapestry.
The Industrial AI Federation is a global coalition of the companies and institutions building industrial AI in production — from the fabs and factories to the banks that finance them and the universities that train their engineers.
The industrial capital of the world. At Kamiyacho Trust Tower — in the heart of the city's business and government district.
A city that has spent a century proving that prestige and precision are the same word.
The fourth edition moves from Stanford to Tokyo — to the source of the industrial discipline this conference is about. Sessions, meals, and the closing reception will be held at Trust City Conference Kamiyacho, directly above Kamiyacho Station and within easy reach of Toranomon, Roppongi, and the Imperial Palace.
IIAC is intentionally small and intentionally senior. A limited number of patronages put your firm in front of the people who decide — the chairs and CTOs of the global industrial-AI giants, from Honeywell and IBM to Panasonic, Tokyo Electron, and JSR, alongside the METI desk, the parliamentarians drafting the next industrial act, and the venture capitalists who fund them all — not in a trade-show hall, but at the table. Tiers and benefits below are draft, to be tailored with each patron.
IIAC is by invitation. Approved requests are confirmed within five business days, with travel guidance and the venue address. Early-bird rates below end August 8, 2026.
Regular $995
Full-day plenary, hosted luncheon, and reception. For practicing engineers, technical leads, and industry attendees.
RequestRegular $2,495
Everything in Professional, plus the Leadership Dinner Reception and eligibility for a curated three-minute lightning talk to the room. Reserved for chair, CEO, and cabinet-level attendees.
RequestRegular $495
Subsidized rate for verified faculty, researchers, and government officials. Limited places.
RequestRegular $395
For verified attendees from low- and middle-income economies — so the global sovereign-AI conversation is actually global. Eligibility by billing address or institutional email.
RequestCancellations refunded in full before September 1, 2026; 50% through September 22; non-refundable thereafter. Registrations may be transferred to a qualified colleague at any time. Attendance is capped to preserve the character of the room. If your firm is sending more than one attendee, please request a patronage.