Enterprise AI

Cohere and Hanwha Bring Generative AI Into Shipbuilding

On January 26, 2026, Hanwha Ocean, Hanwha Systems, and Cohere signed a tripartite MOU in Toronto to build enterprise LLMs and multimodal models for submarine design, production, and operations.

Cohere and Hanwha Bring Generative AI Into Shipbuilding — article cover

On January 26, 2026, local time, at a Korea-Canada industrial cooperation forum in Toronto, South Korea’s Hanwha group signed memoranda of understanding with five Canadian companies. The one that matters most to the AI industry is the tripartite agreement between Hanwha Ocean, Hanwha Systems, and the Canadian AI company Cohere: the three will jointly develop AI — built on Cohere’s enterprise large language models and multimodal models — for submarine design, production planning, and manufacturing.

This is not a concept-demo press release. Behind it sits Canada’s Patrol Submarine Project (CPSP), a procurement race where Korean and German shipbuilding camps face off — and for the first time, AI capability is explicitly part of the bidding coalition’s pitch. For Cohere, it is a flagship case of its “secure enterprise AI” route breaking into a defense supply chain.

One Day in Toronto: Five MOUs

The forum drew a South Korean government and industry delegation, and Hanwha signed five MOUs in total:

  • Hanwha Ocean and Algoma Steel: a new steel mill in Canada, steel supply for next-generation submarines, and new maintenance, repair, and overhaul capacity — with Hanwha Ocean planning to invest up to $250 million
  • Hanwha Ocean, Hanwha Systems, and Cohere: jointly developed AI models for submarine production planning, design, and manufacturing
  • Hanwha Systems and Telesat: cooperation on low earth orbit satellite networks; Telesat aims to launch 198 LEO satellites by year’s end and will develop a new military LEO communications system for Korea
  • Hanwha Systems and MDA Space: synergies around MDA’s Aurora software-defined satellite platform
  • Hanwha Systems and PV Labs: electro-optical and infrared sensor technologies for future defense programs

The shared logic across the five deals is Canada’s “Buy Canadian” procurement climate: bundle technology, investment, and jobs into the bid.

Cohere’s Role: Models, Multimodal, and an Agent Platform

The division of labor is clean: Hanwha Ocean brings shipbuilding domain expertise, Hanwha Systems handles defense systems integration, and Cohere supplies secure enterprise AI.

The technical scope is more concrete than you might expect. The jointly developed models cover both large language models and large multimodal models, aimed at submarine operations and shipyard efficiency at once. Hanwha Systems CEO Jae-il Son framed the effort as spanning “design, production, and operations” for the shipbuilding industry. Andrew Chang, Cohere’s APAC vice president, said the partners will integrate an “agentic AI platform and enterprise-grade AI models,” customize them for shipbuilding and defense, and deploy them into mission-critical applications including the CPSP — citing real-time analysis of operational data and equipment diagnostics in deep-sea operations. Canadian defence outlet Vanguard Defence reports that the pilot will use enterprise-grade generative AI to generate engineering specifications from existing vessel data and new design inputs, with the goal of automating and validating specification documents.

Cohere’s profile makes the deal make sense. Co-headquartered in Toronto and San Francisco, valued at more than $7 billion according to The Korea Times, with investors including Nvidia and Oracle, the company has spent years pitching security-controlled, privately deployable enterprise models — precisely the only kind of AI a defense-adjacent, high-sensitivity industry will seriously consider.

The Real Prize: Canada’s Patrol Submarine Project

The CPSP exists to replace the Royal Canadian Navy’s aging Victoria-class fleet. Hanwha is bidding its KSS-III boat on Arctic deployability and endurance, with a delivery promise attached: four submarines before 2035 if the contract lands in 2026, and the full twelve-boat fleet delivered by 2043. The rival camp is German. Sung-chul Eoh, president of Hanwha Ocean’s Naval Ship Division, called the agreements “a declaration of Hanwha’s commitment to becoming deeply integrated into Canada’s technology ecosystem.”

Seen as bid strategy, the AI partnership is the newest chip on the table: the steel mill covers the supply chain, satellites cover communications, sensors cover mission systems — and Cohere covers the politically useful slot of “Canada’s own AI company.”

Vertical AI’s Next Stop: High-Value Hardware

For AI watchers, the signal here is bigger than one contract. First, enterprise AI competition is shifting from generic assistants to verticals, and shipbuilding — document-heavy, labor-expensive, with long validation cycles — is exactly where LLM ROI is easiest to prove; automating and validating engineering specifications is a textbook high-value document workflow. Second, security and sovereignty claims are the ticket into regulated industries: Cohere’s private-deployment route differentiates it from cloud-first rivals in defense procurement. Third, the path for agent platforms into physical industries is taking shape: from design documents to production planning to equipment diagnostics, shipbuilding offers far more measurable deployments than consumer apps.

Whether it works will show in two places: who wins the CPSP in 2026, and whether the pilot actually gets the specification workflow running.

Sources

AI-assisted summary compiled from the sources above, reviewed by a human before publishing.

FOUND_THIS_USEFUL?

Support more practical AI articles, tutorials, and build notes.

BUY_ME_A_COFFEE
SHAREXEMAIL