Deep-Sea Mining in 2026: TMC, the ISA Debate, and the AI Subsea Stack

Deep-sea mining sits in a uniquely contested place in 2026. The technical case for accessing the polymetallic nodules of the Clarion-Clipperton Zone, the seafloor massive sulphides around hydrothermal vents, and the cobalt-rich ferromanganese crusts on undersea mountains is grounded in the genuine global shortage of nickel, cobalt, manganese, and copper that the energy transition needs. The environmental case against is grounded in the equally genuine reality that deep-sea ecosystems are slow-growing, poorly understood, and extremely sensitive to physical disturbance. The legal status — the International Seabed Authority moratorium debate, the contrasting national positions, the sponsoring-state dynamics — has held the industry in a strange semi-paralysed state where the technology has matured faster than the licensing regime that would allow commercial production.

This post walks through where deep-sea mining sits in 2026 — TMC, GSR, Loke, the ISA debate, the three resource types, and the autonomous subsea vehicle and AI stack that has matured alongside the mining technology.

The Metals Company (TMC) and the sovereign-sponsor move#

The Metals Company, the Canadian-headquartered company most closely associated with the deep-sea mining push, went public via SPAC in 2021 and has been the lead test case for the ISA exploration-to-exploitation transition. TMC holds exploration contracts through the sponsoring states of Nauru, Tonga, and Kiribati covering blocks of the Clarion-Clipperton Zone in the central Pacific. In 2022 and 2023, TMC and its partner Allseas operated the prototype Hidden Gem mining vessel for the first integrated deep-sea nodule collection test in the CCZ, recovering several thousand tonnes of nodules from the seabed.

The 2024 and 2025 period was dominated by TMC’s legal manoeuvres around the ISA application process. In June 2024 Nauru formally invoked the “two-year rule” in the UN Convention on the Law of the Sea, which triggers an obligation for the ISA to either finalise commercial-mining regulations or process applications under provisional rules. Through 2025 TMC pushed against the slow pace of the ISA mining-code negotiations, and in early 2026 the company filed for permits under a contested interpretation of US domestic deep-sea mining law (the 1980 Deep Seabed Hard Mineral Resources Act) — an end-run around the ISA process that has been controversial in international policy circles and has been challenged on multiple grounds.

The TMC strategy in 2026 sits in tension with the broader international consensus on how seabed mining should be regulated, and the legal outcome of the US-permit path is genuinely uncertain.

GSR, Loke, and the other operators#

Global Sea Mineral Resources (GSR), a subsidiary of the Belgian dredging conglomerate DEME, has been the more technically conservative of the major players. GSR operates the Patania II nodule collector prototype, which conducted full-scale CCZ trials in 2021 and follow-on tests through 2023. GSR’s positioning has emphasised the regulatory and environmental-baseline work, and the company has been a participant in the European-Commission-funded research programs that aim at scientifically rigorous environmental impact assessment.

Loke Marine Minerals, the Norwegian company, focuses on the Norwegian continental-shelf opportunity. The Norwegian parliament’s January 2024 vote to open Norwegian Sea areas to seabed-mining exploration was a major political milestone for the industry, although a December 2024 reversal — driven by environmental opposition inside the governing coalition — pulled the immediate licensing rounds. Loke continues operating, and the longer-term Norwegian regulatory status remains uncertain.

Other operators in the field include Impossible Metals (which is pursuing a more selective nodule-collection approach using AI-driven robotic arms rather than vacuum collectors), Ocean Minerals LLC, and a small number of Chinese and Russian state-affiliated operators that hold ISA exploration contracts but have been less commercially active in the public eye.

The ISA moratorium debate#

The International Seabed Authority, the UN body that administers seabed-mining outside national waters, has been the centre of the regulatory debate. As of mid-2026, the ISA has not finalised the commercial-mining regulations (the “Mining Code”) despite years of negotiation. The country positions have polarised. France, Germany, Costa Rica, Chile, Palau, New Zealand, and a growing list of others have called for a moratorium or precautionary pause on commercial seabed mining. China, Russia, India, South Korea, and some smaller sponsoring states have pushed for the regulations to be finalised so that commercial operations can begin.

The 2024 and 2025 ISA Council sessions advanced parts of the Mining Code text but did not finalise the regulations. The 2026 sessions have continued the negotiations under significantly more public scrutiny than the ISA process has historically faced, with environmental NGOs (including Greenpeace, the Deep Sea Conservation Coalition, and WWF) maintaining sustained pressure for a moratorium.

The honest reading of the regulatory status in 2026 is that the ISA has not authorised commercial mining, the political consensus required to do so does not exist, and the industry’s path to actual commercial production runs through some combination of further ISA process, national-waters projects (the Norwegian-shelf model), and the contested US-domestic-law interpretation that TMC has invoked.

The three resource types — nodules, vents, crusts#

Three deep-sea resource types dominate the commercial conversation. Polymetallic nodules — the potato-sized iron-and-manganese concretions that lie loose on the abyssal sea floor in the Clarion-Clipperton Zone and a handful of other regions — are the easiest to access (no drilling required, just vacuum collection or robotic pickup) and the focus of TMC, GSR, and the other CCZ operators. The metal content is dominated by nickel, copper, cobalt, and manganese.

Seafloor massive sulphides, which form around hydrothermal vents on mid-ocean ridges and back-arc basins, are richer in copper, zinc, gold, and silver but much more environmentally contentious because hydrothermal vents host unique chemosynthetic ecosystems. Cobalt-rich ferromanganese crusts form on the flanks of submarine mountains and contain cobalt, platinum, tellurium, and rare-earth elements. Crusts are harder to extract because they require breaking and grinding rock rather than picking up loose nodules.

The commercial focus has settled overwhelmingly on nodules. The technical access is easier, the resource is more geographically concentrated and better-characterised, and the environmental footprint of nodule collection — while still significant — is somewhat less acute than the vent or crust alternatives.

Autonomous subsea vehicles and the AI for resource assessment#

The technology stack underneath deep-sea mining has matured alongside the mining-vehicle work and is independently interesting. Autonomous underwater vehicles (AUVs) from Kongsberg, Saab Seaeye, ECA Group, and the newer entrants like Bedrock Ocean Exploration and Terradepth operate at depths of 3,000 to 6,000 metres for resource-survey and environmental-monitoring missions. The vehicles use multibeam sonar, sub-bottom profilers, magnetometers, and high-resolution optical imaging to characterise the seafloor.

The AI layer on top of this data has become the differentiating technology. Computer-vision models trained on millions of nodule and seabed-fauna images do the resource assessment and the environmental baseline characterisation that used to require many person-years of manual image labelling by marine biologists. The same models support the operational decisions during mining — selectively picking up nodules while leaving surrounding fauna undisturbed, in the Impossible Metals approach — and the post-disturbance environmental monitoring that any commercial operation will be required to conduct.

The broader deep-tech adjacencies — high-bandwidth subsea communications, underwater inertial navigation, long-endurance battery and energy systems for remote operation — sit at the intersection of deep-sea mining, offshore wind installation, subsea oil-and-gas decommissioning, and naval applications. The same Kongsberg or Saab AUV that surveys a CCZ mining block also inspects offshore wind cables and performs naval mine-countermeasure work.

What this means in 2026#

The realistic 2026 read on deep-sea mining is that the technology is mature enough for commercial operation, the resource is real and meaningful for the critical-minerals shortage, the environmental concerns are serious and unresolved, and the regulatory path remains contested. Whether seabed mining becomes a significant component of the global critical-minerals supply by the early 2030s, or remains permanently constrained by the ISA moratorium dynamics, is one of the most consequential open questions in the broader energy-transition supply chain.

Where pdpspectra fits#

Our data-engineering practice and AI and LLM integration practice build the high-throughput marine-data and computer-vision platforms that subsea-survey operators, environmental-monitoring programs, and resource-assessment teams need.

Related reading: the AI energy utilities post, the lunar economy 2026 post, and the edge AI deployment patterns post.

---

Deep-sea mining is technically ready and politically contested; the next few years decide which way it goes. Talk to our team about your subsea-data program.