Anduril launches Copperhead AUV and Seabed Sentry at world's largest naval trade show

Anduril used the Sea-Air-Space conference in Washington — the world's largest naval trade show — to launch two new maritime defense products: Copperhead, a high-speed autonomous underwater vehicle with a torpedo-capable munitions variant, and Seabed Sentry, a cableless wireless undersea surveillance system. Shane Arnott, Anduril's SVP for maritime, walked through both products and the strategic logic behind them.

Seabed Sentry

The starting point is a surveillance gap that Arnott says is broadly underappreciated: roughly 70–80% of the seafloor remains uncharted, and the infrastructure running beneath it — subsea cables carrying 90%+ of global internet traffic, gas pipelines, energy links between nations — is largely unmonitored. Existing undersea surveillance systems are Cold War-era designs that require large cabling ships to deploy, are nearly impossible to upgrade, and leave their positions visible to adversaries.

Seabed Sentry is designed around the opposite logic. It's fully wireless and fully electric, which means units can be placed covertly anywhere on the ocean floor, including via one of Anduril's autonomous submarines. Arnott sees the customer base as dual-use — defense and military allies, but also energy infrastructure operators. The system also monitors biological activity, specifically so Anduril can tune down acoustic output when marine mammals are passing through a surveillance area.

“Seabed Sentry is about undersea awareness. Some 70 to 80% of the seafloor is uncharted. 90% plus of internet traffic runs across subsea cables — many nations have their gas lines under the waves. These are broadly unprotected, unwatched right now. Copperhead is built to be at a fraction of the cost of a mark 48 or mark 54. We can build them at thousands per year, whereas right now you may be lucky to get 200 torpedoes out of a supply chain as it exists today.”

Copperhead

Copperhead is a fast, software-defined AUV with a munitions variant — effectively a torpedo — engineered to address what Arnott frames as a broken unit economics problem in undersea warfare. Current torpedoes like the Mark 48 and Mark 54 cost far more than the unmanned underwater vehicles proliferating from China and Russia. Launching a Mark 48 against a cheap UUV is, in Arnott's words, a 10-to-1 cost mismatch that doesn't work at the scale of a conflict in the Asia-Pacific or the High North.

Copperhead is built to be a fraction of the cost of those legacy weapons and manufacturable at thousands of units per year. Arnott says current U.S. torpedo supply chains struggle to produce around 200 units annually. The system is also fully electric, which matters for operations under Arctic ice — an environment Arnott flags as increasingly contested as ice caps retract and nations compete for High North resources.

The rectangular form factor is a deliberate choice. Arnott says Anduril's founder Palmer Luckey pushed for a square design, and the engineering analysis confirmed it actually produces faster, more stable flight for their use case. It's also significantly easier to manufacture than a cylindrical pressure vessel. While the diameter is similar to existing torpedo tubes, Arnott says the current intent is for Copperhead to be launched from Anduril's own autonomous submarines rather than integrated into legacy platforms — though customer interest in legacy compatibility is prompting Anduril to explore it.

The system-of-systems pitch

At Sea-Air-Space, Anduril is demonstrating how the pieces fit together. Arnott says the company's large autonomous submarine, the Dive XL, can carry dozens of Copperheads and a dozen Seabed Sentry units simultaneously. That mix-and-match capability — mothership plus surveillance plus munitions — isn't immediately intuitive, so Anduril is running virtual reality demos alongside the hardware to give naval commanders a sense of how they'd actually employ the system.

Why autonomy matters most underwater

Arnott is direct that the subsea domain is harder than space, air, surface, or land for autonomous systems. There's no GPS. Communications pipes are either extremely thin or nonexistent. Vehicles must operate on pre-planned missions and share only tiny amounts of information with other assets. That constraint has forced Anduril to build what Arnott describes as its most capable autonomy engineering team anywhere in the company.

The complexity compounds when kinetic action is involved. Arnott says the core challenge isn't building an autonomous submarine — that part, he argues, is tractable. The challenge is ensuring the vehicle can survive, communicate, and still keep a human in the approval chain when weapons are being deployed. Being both the platform developer and the weapons developer means Anduril can iterate on seekers, threat profiles, and safety cases faster than a contractor integrating someone else's effector.

Takeaway: Anduril is making an explicit bet that the next major warfare domain is undersea, and that legacy torpedo economics and Cold War surveillance infrastructure leave a wide opening. Copperhead and Seabed Sentry are designed to be cheap enough to deploy at scale, covert enough to evade detection, and electric enough to operate under Arctic ice — the three constraints Arnott says current systems fail on simultaneously.