SoftBank Commits to Major AI Data-center Investment

SoftBank CEO Masayoshi Son told CNBC on June 1 that the AI revolution is "50x bigger" than the dotcom boom, calling it "the biggest revolution of technology," CNBC reports. CNBC reports the comments came as SoftBank announced investments to build AI infrastructure in France, including a cited target of 5 GW of AI data-center capacity and a reported plan to build 3.1 GW of AI data centers in the northern Hauts-de-France region by 2031, with sites named in Dunkirk, Bosquel, and Bouchain. CNBC reports Son made the remarks at a press briefing attended by French President Emmanuel Macron and that SoftBank is partnering with a French engineering company to create an industrial production hub in Dunkirk. CNBC also quotes Son referencing the Wall Street crash of 1929 and saying "there's always a correction."

CNBC provides the capacity figures as part of SoftBank's announcement: a broader 5 GW target was cited in coverage, and a more-specific 3.1 GW buildout is reported for the Hauts-de-France sites by 2031. CNBC reports the program includes multiple named locations and a partnership with a French engineering firm; the scraped CNBC article does not supply the engineering firm's name in the text provided here.

Editorial analysis: Large-scale AI data-center deployments of the size reported typically require sustained capital commitments, long-term power contracts, and coordination with local authorities and grid operators. Industry observers note that securing multi-gigawatt capacity often involves staged construction, industrial partners for physical infrastructure, and local regulatory engagement. For practitioners, increasing regional capacity tends to reduce model training and inference latency domestically and can change where cloud and enterprise customers procure large-scale GPU/accelerator capacity.

Editorial analysis: Observers should track:

• •permitting and grid-connection timelines for the named Hauts-de-France sites
• •announcements naming the engineering partner and details on power sourcing or dedicated generation
• •any announcements from regional utilities or French regulators on capacity reservations or transmission upgrades. These signals will clarify how soon multi-gigawatt capacity might come online and what segments of compute (training, inference, hyperscale) it will serve

Editorial analysis: The combination of a high-profile investor statement and a multi-gigawatt infrastructure commitment highlights continued private capital interest in on-premises and regional AI compute. For ML teams and infrastructure planners, regional capacity additions can shift considerations for data residency, procurement timelines, and hybrid-cloud architectures.