On June 26, 2026, Japan’s NTT detailed the creation of the IOWN AI Fund, targeting around ¥80 billion to invest in photonics, semiconductors and AI-related startups that fit its IOWN next-generation network vision. The fund, formed after a June 10 announcement, will be managed with partners including Walden Catalyst Ventures and will invest in later-stage companies while Walden backs earlier-stage firms. Telecom and chip players such as SK Telecom, SK hynix and Chunghwa Telecom are participating as partners.
This article aggregates reporting from 2 news sources. The TL;DR is AI-generated from original reporting. Race to AGI's analysis provides editorial context on implications for AGI development.
NTT’s IOWN AI Fund is Japan’s bid to make sure it owns a strategic slice of the AI compute stack rather than just renting GPUs from US hyperscalers. By tying capital to its photonics‑heavy IOWN roadmap, NTT is trying to do for low‑power, optically‑interconnected infrastructure what NVIDIA did for GPU‑centric AI: build a tight ecosystem in which hardware, network and software co‑evolve. The presence of SK Telecom, SK hynix and Chunghwa Telecom underscores that this is a regional industrial policy move as much as a corporate one.
In the AGI race, this matters because future frontier systems will be constrained by power and interconnect more than by raw FLOPs on a single die. If IOWN‑style optical networks and memory architectures deliver meaningful efficiency gains, they could enable smaller players to run serious models without matching the hyperscalers’ power budgets. Conversely, if the ecosystem fragments across incompatible stacks, it could slow standardization and make it harder for labs to target a single hardware abstraction.
Either way, the creation of a dedicated AI fund anchored in network and photonics technology shows that “compute sovereignty” is now a capital‑allocation thesis in its own right, not just a talking point in government strategy papers.