Who Owns the Cables — Hyperscalers, State Champions, and the Battle for Submarine Infrastructure
Information Infrastructure — The Physical Internet | Part 4 of 8
Opening — Who Actually Owns the Internet’s Backbone
The submarine cable system that carries roughly 99% of intercontinental data traffic is not owned by the governments whose economies depend on it. (Fact) It is owned, in shifting proportions, by four kinds of actors: a shrinking tier of legacy telecom carriers, an expanding tier of American hyperscalers, a small set of cable-manufacturing firms whose ownership structures are themselves becoming geopolitical instruments, and — newest in the mix — Chinese state-linked entities operating through both pure-Chinese builds and minority positions in Western-led consortia.
This is not the architecture the system was designed to have. The post-2000 model assumed cables would be neutral commercial infrastructure built by competing private vendors, financed by telco consortia, and laid wherever the bandwidth economics worked. (Assessment) What replaced that model — over roughly fifteen years — was a structure in which the buyers of capacity became the owners of cables, in which the manufacturing base bifurcated along US-China lines, and in which two governments (Washington and Paris) have intervened directly in the ownership of cable companies for national-security reasons. The first piece of that response was a US executive order. The second was a French acquisition. Neither is unwinding.
This article maps that ownership transition, profiles the four major cable manufacturers, dissects the Digital Silk Road cable-participation pattern, and reads the institutional response — beginning with the Pacific Light Cable Network (PLCN) decision — as the template for how Western governments now police submarine-cable access. Earlier parts of this series mapped the physical layer and its chokepoints. This part is about who controls it.
The Old Model — Telco Consortia and Why It Changed
For roughly half a century after the first transatlantic telephone cable (TAT-1, 1956), submarine cables were built by consortia of national telephone monopolies. (Fact) AT&T, British Telecom, France Télécom, NTT, Telstra, and the equivalents in other landing countries shared construction costs, allocated capacity by ownership share, and operated cables as common-carrier infrastructure. The model worked because international voice traffic was the dominant use case, the buyers and the carriers were the same actors, and capacity demand grew predictably.
Two shifts broke that model. (Fact) First, telephony stopped being the dominant traffic and was replaced by IP data — which grew exponentially rather than linearly and which originated overwhelmingly from a handful of content companies. Second, deregulation in the 1990s and 2000s fragmented the legacy carriers into competing entities whose individual balance sheets could no longer carry the multi-billion-dollar capex for new cable systems.
(Assessment) By the late 2000s, the legacy telco consortia could finance maintenance and capacity upgrades on existing routes but could not, on their own, build the new generation of high-fiber-count systems that the cloud era was about to require. The buyers of capacity were going to have to become the builders. They did.
The Hyperscaler Takeover — <10% to 66%+ in One Decade
Before 2012, content providers — Google, Meta (then Facebook), Amazon Web Services, and Microsoft — collectively held less than 10% of global submarine cable capacity. (Fact, TeleGeography) By 2020 they held approximately 66%. On the trans-Pacific route the share is roughly 78%; on intra-Asia routes roughly 75% (2022–2023 data, TeleGeography). The count of cables in which content providers hold investor or owner positions rose from 20 in 2017 to 59 in the most recent TeleGeography tally — a roughly 3× expansion in about seven years.
The inflection point is conventionally placed around 2010–2012, with Google’s Unity cable investment as the early signal. The major acceleration ran from 2016 to 2020. (Fact) TeleGeography’s pipeline of announced new cable builds for the 2024–2026 period totals approximately $11 billion, with internet giants accounting for the bulk of that capital.
What drove the shift was a combination of three factors. (Assessment) First, the economics of cloud regions made bandwidth a strategic input, not a commodity purchase: a delay of weeks in lighting a new region cost more than the marginal cost of owning the cable. Second, hyperscalers achieved capital scale that no national telco could match — the four firms combined hold cash and short-term investments measured in hundreds of billions of dollars, and submarine cables at roughly $300–500 million per system are a rounding error against that base. Third, owning the cable allowed end-to-end optimization (custom fiber pairs, software-defined wavelength allocation, open-line systems) that consortium cables could not match.
The strategic consequence is that four US-headquartered private companies now collectively own a majority of the world’s submarine capacity. (Assessment) They are not states. They are not common carriers in the legal sense that legacy telcos were. They are not subject to the same regulatory regime that governed the cable system for its first fifty years. The infrastructure layer of the internet is now corporate property in a way that is historically novel — and it is concentrated in a small number of US legal jurisdictions, which is itself a geopolitical fact.
The Four Manufacturers — and What Their Ownership Means
Beneath the buyer layer sits the manufacturing base. Four firms build essentially all the world’s submarine cable systems: SubCom (United States), Alcatel Submarine Networks / ASN (France), NEC Networks (Japan), and HMN Technologies (China). The ownership trajectory of each one over the past decade tells the geopolitical story more cleanly than any policy document.
SubCom — Cerberus, Continuation, and the US Anchor
SubCom traces its lineage to AT&T’s marine division, founded in 1955. (Fact) The corporate path runs: AT&T marine division (1955) → sold to Tyco International (1997) → Tyco Telecommunications → TE Connectivity → acquired by Cerberus Capital Management on November 5, 2018, for $325 million → renamed SubCom LLC. (A common error in earlier reporting placed SubCom under General Dynamics ownership at some stage. That is incorrect; General Dynamics has not held SubCom.) (Fact) In April 2026, Cerberus closed a $2.3 billion continuation vehicle for SubCom, extending the ownership commitment and signaling that the US private-equity owner intends to hold the asset across the current build cycle.
SubCom is the first company to deploy a cumulative 1 million km of subsea cable. (Fact) Its recent project list — SEA-ME-WE-6, Google’s Grace Hopper and Firmina, the six-member JUPITER trans-Pacific system, multiple Indo-Pacific systems — places it as the preferred US-government-acceptable builder for Western hyperscaler projects. (Assessment) Cerberus ownership is not formally a national-security designation, but it is a US-domiciled private-equity sponsor with deep defense-industrial connections, and the firm’s project portfolio increasingly tracks the routes where Team Telecom approvals matter.
Alcatel Submarine Networks — Now French State
ASN’s trajectory is the most consequential ownership change in the industry over the past five years. (Fact) Its lineage runs from Alcatel-Lucent through the 2016 Nokia acquisition. On December 31, 2024, Nokia closed the sale of an 80% stake in ASN to the French state, via the Agence des participations de l’État (APE). Nokia retains 20% and board representation. ASN is now a French-state-majority entity.
ASN operates seven cable-laying vessels and has deployed more than 750,000 km of cable in its lifetime; it claims the world No. 1 position by installed base. (Fact) Its current flagship project is the 2Africa system, the 45,000 km cable that rings the African continent. (Assessment) The French state’s decision to nationalize a Western cable manufacturer is the clearest signal yet that European governments now treat submarine cable manufacturing capacity as strategic infrastructure on the same tier as semiconductor fabs or nuclear energy — not a commercial sector to be left to market consolidation.
NEC Networks — The Japanese System Integrator
NEC Corporation (TSE: 6701) is the system integrator; OCC Corporation — a subsidiary jointly held by NEC (~75%) and Sumitomo Electric (~25%) following NEC’s 2008 acquisition of the controlling stake — manufactures the optical fiber cable. (Fact) NEC has deployed more than 400,000 km of cable since 1964 and recently built Google’s Echo cable (US–Singapore) and the Asia Direct Cable (ADC, completed December 2024, approximately 10,000 km, 160 Tbps).
A key structural limitation: NEC does not own cable-laying vessels. (Fact) It contracts laying ships from third parties on a per-project basis. This makes NEC more flexible in some respects but vulnerable to capacity constraints when the global ship pool is overcommitted — a recurring problem as build volumes accelerate.
HMN Technologies — The Chinese Entrant
The fourth manufacturer is the geopolitical hinge. HMN Technologies — formerly Huawei Marine Networks — is the only Chinese firm at the manufacturing tier, and its existence is the proximate cause of every Western institutional response described later in this article.
HMN Technologies — The Chinese State-Linked Manufacturer in Detail
HMN Technologies’ corporate history is short but compressed with geopolitical signal. (Fact) The firm was established in 2008 as a joint venture between Huawei Technologies (51%) and Global Marine Systems of the UK (49%) under the name Huawei Marine Networks. In May 2019, Huawei Marine was added to the US Entity List as part of the broader Huawei sanctions regime. In 2020, Hengtong Optic-Electric Co. Ltd. (Shanghai-listed, SZSE: 600487) acquired Huawei Technologies’ 81% stake; the company rebranded as HMN Technologies Co. Ltd. on November 3, 2020. On March 6, 2023, Hengtong acquired the remaining 19% from Global Marine Holdings LLC for $32 million, completing the transition. HMN Tech is now a 100% Hengtong subsidiary.
The ownership structure of Hengtong itself is publicly available and instructive. (Fact) The largest single shareholder is founder Genliang Cui, holding approximately 24%; retail investors hold approximately 52%. On paper, this is not a state-owned enterprise. (Assessment) In practice, the firm receives Chinese state subsidies, and HMN bids in international cable tenders have been reported by Nikkei and the Atlantic Council to come in “around a third cheaper than competing proposals” — a margin that competitive analysts attribute primarily to subsidy support rather than productivity advantage.
The structural access risk is independent of any specific exploitation evidence. (Fact) Article 7 of China’s 2017 National Intelligence Law obligates all Chinese organizations and citizens to “support, assist, and cooperate with national intelligence work in accordance with the law.” This applies to HMN Technologies as a Chinese corporate entity regardless of nominal private-shareholder status. (Assessment) The risk model that Western governments operate from is not “HMN has been caught tapping cables” — there is no public evidence to that effect — but rather “HMN is structurally compelled to cooperate with PRC intelligence services if asked, and a manufacturer with deep technical access to a cable system has capability that no end-user audit can fully verify.”
HMN’s project portfolio per the Submarine Cable Almanac 2023 reaches 16 cable projects worth approximately $1.6 billion across 27 countries. (Fact) The flagship is the PEACE Cable (Pakistan East Africa Connecting Europe), framed explicitly as a Belt and Road project. Other projects include the Asia Link Cable System and multiple Africa–Middle East–Pacific routes. The geographic pattern is consistent: HMN dominates where Western consortia have either been priced out or where landing-country governments have a political relationship with Beijing that overrides Western pressure.
The ASN Surprise — Why France Just Bought a Submarine Cable Company
The French state’s December 2024 acquisition of 80% of ASN deserves separate treatment because it is the clearest signal that the Western response to the Chinese manufacturing entry has moved beyond export controls into industrial-policy ownership. (Fact)
Nokia, a Finnish telecom-equipment firm, had inherited ASN through the 2016 Alcatel-Lucent acquisition and never integrated it strategically. The submarine cable business runs on multi-year project cycles, low margins relative to telecom-equipment leasing, and capital intensity that does not match a software-driven telecom-equipment vendor’s portfolio. (Assessment) Nokia’s commercial logic for divesting ASN was sound. The geopolitical logic for who acquired it is what matters.
The French state did not allow ASN to be sold to a private-equity sponsor or to a non-European industrial buyer. APE acquired the 80% directly. (Fact) France now owns the largest Western submarine cable manufacturer by installed base, with vessels, cable plants in Calais, and a project pipeline that includes 2Africa and multiple Atlantic routes. (Assessment) The acquisition implies three policy positions: (1) Paris assesses submarine cable manufacturing as strategic infrastructure that cannot be allowed to consolidate into Chinese hands; (2) the EU-level industrial-policy machinery did not move fast enough to nationalize-via-coalition, so France acted unilaterally; (3) ASN will now be available to French and EU strategic projects on terms that a private owner would not have offered.
The closest analogue in recent history is the French state’s earlier interventions in Areva (nuclear) and Alstom (rail). (Assessment) Cables now sit in that category.
US Team Telecom — The Institutional Response
The American response is institutional rather than ownership-driven. The vehicle is Team Telecom — formally, the Committee for the Assessment of Foreign Participation in the United States Telecommunications Services Sector — established by Executive Order 13913, signed April 4, 2020. (Fact) The committee is composed of the Department of Defense, the Department of Justice, and the Department of Homeland Security, and it advises the Federal Communications Commission on national-security implications of foreign participation in US telecom licenses, including submarine cable landing licenses.
Team Telecom’s powers are concrete. (Fact) It can recommend that the FCC deny a license application, grant it with conditions (typically a network-security agreement and audit rights), modify an existing license, or revoke one. The FCC has, in practice, followed Team Telecom recommendations in every cable case since EO 13913.
The PLCN Case — Template Decision
The Pacific Light Cable Network is the landmark precedent. (Fact) PLCN was a Google–Meta-led trans-Pacific cable originally designed to land in the United States, Hong Kong, Taiwan, and the Philippines. In June 2020, Team Telecom recommended that the FCC deny specifically the US–Hong Kong segment, citing China Telecom Americas’ state-affiliated investor status in the cable and the PRC laws compelling Chinese entities to hand over data to intelligence services on request.
Google and Meta withdrew the Hong Kong landing application and refiled for the US–Taiwan and US–Philippines segments only. (Fact) In January 2022, the FCC granted landing licenses for those segments. The cable is operational today on its approved routes.
The precision of the decision matters analytically. (Assessment) Team Telecom did not block the cable. It blocked the Hong Kong segment, surgically removing the PRC-jurisdiction landing point while permitting the rest of the system to operate. The signal to industry was unambiguous: cables with Chinese state-affiliated investor participation or PRC landing points face structural blocks; cables that route the same capacity through US-aligned landing points (Taiwan, the Philippines, Japan, Singapore, Guam) get through. The pattern has held in every subsequent review.
Since PLCN, Team Telecom has reviewed and imposed conditions on every cable system with significant Chinese corporate participation seeking US landings. (Fact) The downstream effect is that hyperscalers now route around China at the planning stage: cable systems are designed with no PRC landing point and no Chinese consortium members before the FCC application is filed, because the cost of restructuring a cable after Team Telecom review is prohibitive.
The Digital Silk Road — How Chinese State Participation Is Structured
The Digital Silk Road — the digital-infrastructure component of the Belt and Road Initiative — first appeared in formal Chinese policy in the 2015 NDRC document “Vision and Actions on Jointly Building Silk Road Economic Belt and 21st-Century Maritime Silk Road.” (Fact) Of the document’s 5,598 words, only 56 address digital and cable infrastructure, referring to an “International Information Silk Road” that includes “transcontinental submarine optical cable projects.” That brevity is itself analytically interesting — the digital component was, in 2015, a junior partner to road, rail, and port. By 2020 it had become the most strategically contested element.
Chinese state participation in the cable layer operates through two distinct channels. (Assessment)
Channel 1: Pure-Chinese builds. The PEACE Cable is the archetype. (Fact) It is 100% Chinese corporate ownership, built by HMN Tech, landing at Gwadar (a China-Pakistan Economic Corridor port), and described by the Atlantic Council as “funded, owned, and constructed entirely by Chinese entities.” These are unambiguously instruments of Belt and Road Initiative policy. They do not face Team Telecom review because they do not seek US landings.
Channel 2: Minority equity participation in Western-led consortia. This is the channel that Team Telecom is calibrated against. (Fact) Documented examples include:
- 2Africa: China Mobile International (a state-owned enterprise under SASAC) is one of eight consortium members.
- AAE-1 (Asia-Africa-Europe-1): China Unicom (majority state-owned, SASAC) is a consortium member.
- SEA-ME-WE-6: China Unicom holds a non-landing equity position.
- Asia Direct Cable (ADC): China Mobile International and China Telecom Global both participated.
The pattern is structural. (Assessment) Chinese state-owned carriers buy minority equity in cables they do not control, gaining (a) commercial capacity at cost, (b) technical visibility into the system’s design, and (c) seats at consortium governance tables. None of these positions individually triggers a national-security review threshold in most jurisdictions. Collectively, they give Beijing distributed presence across most major non-US-only cable systems built since roughly 2015.
A particular landing-country pattern reinforces the geometry: Djibouti Telecom (100% government-of-Djibouti owned) is the landing party for DARE1, AAE-1, EASSy, and PEACE Cable. (Fact) Djibouti also hosts China’s first overseas military base, established in 2017. The colocation of Chinese-affiliated cable termination and Chinese military basing at the Bab el-Mandeb chokepoint is not a single-cable issue — it is a system-level convergence of digital and physical infrastructure under Digital Silk Road policy.
The Bifurcation — Two Supply Chains as a Structural Feature
What has emerged over the past five years is no longer a single global cable industry with a Chinese participant. (Assessment) It is two parallel supply chains operating on partially overlapping geography.
The Western supply chain — SubCom, ASN, NEC — builds the cables that land in the United States, the EU, Japan, Australia, and US-aligned Indo-Pacific states. These cables are increasingly financed by US hyperscalers, increasingly designed to avoid PRC landing points and Chinese consortium participation, and increasingly reviewed at the planning stage against Team Telecom criteria.
The Chinese supply chain — HMN Technologies, financed by Hengtong, supported by state subsidy, integrated into Belt and Road Initiative capital flows — builds the cables that land in jurisdictions where Western political pressure does not bind: Pakistan, parts of East Africa, Djibouti, smaller Pacific states, and Eurasian Belt-and-Road partners. These cables typically do not seek US landings and do not enter the Team Telecom system at all.
The two systems still touch at landing-country interconnects, at IXPs, and at the four manufacturers’ shared dependence on a small pool of cable-laying vessels and rare-earth materials for repeaters and amplifiers. (Assessment) But the architectural pattern — who builds what, who lands where, who reviews whom — has bifurcated to a degree that would have been hard to predict in 2015 and that no current policy mechanism on either side appears to be reversing.
Strategic Implications
For the United States. Team Telecom is now the de facto gatekeeper of the trans-Pacific and trans-Atlantic cable layer for any system seeking US landings. (Assessment) The mechanism is effective at the application stage but does not address legacy cables, third-country routes that bypass US landings, or the Chinese state-equity positions already embedded in operational consortium cables. The remaining policy gap is upstream: there is no analog to Team Telecom for component-level supply chain (repeaters, optical amplifiers, branching units) where Chinese suppliers retain market presence.
For Europe. France has acted; the EU has not. (Assessment) ASN under French ownership solves the manufacturing-capacity question for France and adjacent EU states but leaves the rest of the continent dependent on a manufacturer that now has a national flag. Brussels will need to decide whether ASN is treated as European strategic infrastructure or as a French industrial-policy asset. The two framings imply different procurement, governance, and crisis-response architectures.
For hyperscalers. Owning cables solves the bandwidth-cost problem but creates a political-exposure problem that did not exist when telcos were the cable buyers. (Assessment) Google, Meta, Amazon, and Microsoft are now infrastructure operators with national-security review obligations in every major jurisdiction where they land cables. The corporate-affairs function at each company has had to absorb a portfolio of state-relations work that legacy telcos used to handle structurally. This is not reversing; it is intensifying.
For Chinese policy. The Digital Silk Road cable program has succeeded at building cables to friendly landing countries and at embedding minority equity positions in non-Chinese consortia. (Assessment) It has not succeeded at securing PRC landings for cables with US connectivity — PLCN was the test case and the test was lost cleanly. The medium-term Chinese response is visible in the data: more pure-Chinese systems, more equity in non-US-landing consortia, more cable diplomacy at landing-country level. None of this requires a doctrinal shift; the existing toolkit suffices.
For analysts. The submarine cable layer is no longer an obscure commercial-infrastructure topic. (Assessment) It now sits at the same analytical tier as semiconductor supply chains and critical minerals — a domain where ownership structure, manufacturing concentration, and state-of-origin determine national exposure. The dataset is small (~550 systems globally), the actor list is short (four manufacturers, ten landing-country categories, four hyperscalers, a handful of Chinese state firms), and the public-source visibility is high. This is one of the more tractable infrastructure-geopolitics problems for OSINT — and one of the higher-stakes ones.
Sources
- TeleGeography, Submarine Cable Map and Industry Reports, 2017–2024. Capacity share and project-count figures. (High confidence)
- Submarine Cable Almanac 2023. HMN Technologies project portfolio. (High confidence)
- Atlantic Council, China’s Digital Silk Road: A Game Changer for Asian Economies and follow-on cable-infrastructure analyses, 2022–2024. PEACE Cable framing, Hengtong subsidy assessment. (High confidence)
- Nikkei Asia, multiple reports 2021–2024 on HMN Technologies bidding patterns and pricing. (High confidence)
- Federal Communications Commission, public dockets for PLCN, SEA-ME-WE-6, and related cable landing license applications, 2019–2024. (High confidence)
- Executive Order 13913, April 4, 2020, Establishing the Committee for the Assessment of Foreign Participation in the United States Telecommunications Services Sector. (Fact)
- NDRC of the People’s Republic of China, Vision and Actions on Jointly Building Silk Road Economic Belt and 21st-Century Maritime Silk Road, March 2015. (Fact)
- PRC National Intelligence Law (2017), Article 7. (Fact)
- Cerberus Capital Management, public announcements re SubCom acquisition (November 2018) and April 2026 continuation vehicle. (High confidence)
- Nokia / Agence des participations de l’État (APE), joint announcements re ASN transaction, closed December 31, 2024. (High confidence)
- Shanghai Stock Exchange filings for Hengtong Optic-Electric Co. Ltd. (SZSE: 600487) shareholder structure. (High confidence)
Part 4 of 8 — Information Infrastructure: The Physical Internet Next: Part 5 — Cable Repair, Resilience, and the Vessel Pool
Related: SYNTHESIS · Digital Sovereignty · Economic Chokepoints — Coercive Statecraft · Five Eyes · GCHQ