The Royal Australian Navy (RAN) just slapped a "Division 9" sticker on its autonomous systems unit. The press releases are glowing. They talk about innovation. They talk about keeping sailors out of harm's way. They talk about a "sovereign capability" that sounds impressive over a morning flat white in Canberra.
It is a lie. Not a malicious one, but a strategic one designed to mask a terrifying reality: Australia has the hulls, but it doesn't have the humans. Meanwhile, you can explore similar stories here: Amazon Globalstar Deal Changes Everything for Your Smartphone.
Establishing a dedicated unit for uncrewed surface vessels (USVs) and underwater drones isn't a bold leap into the future. It is a desperate pivot. We are witnessing the "Uber-fication" of the Pacific—an attempt to replace expensive, unionized, high-maintenance human assets with an algorithmic fleet because the RAN’s recruitment numbers are in a death spiral.
The Ghost Fleet Myth
The industry consensus is that autonomous systems are an "additive" force multiplier. The logic suggests that if you have ten frigates, adding fifty drones makes you five times stronger. To see the complete picture, check out the excellent report by CNET.
This is fundamentally flawed. In the current geopolitical climate, drones aren't being added to the fleet; they are being used to justify the shrinking of the fleet’s human core. Division 9 is a placeholder for the sailors that aren't coming.
I have seen defense departments burn through billions on "innovation hubs" that produce little more than expensive toys for trade shows. The RAN’s move to formalize Division 9 is a signal to the market, but it lacks the industrial backbone to survive a high-intensity conflict.
Autonomous systems require a massive tail of shoreside technicians, data analysts, and cybersecurity experts. You aren't "removing the human." You are just moving the human from a bunk on a ship to a desk in Adelaide. And right now, the ADF is struggling to fill both.
Hardware is Cheap Intelligence is Fragile
The hype focuses on the hardware. We see sleek, carbon-fiber hulls like the Ocius Bluebottle or the Anduril Dive-LD. They look like the future.
But look at the math of modern naval warfare. An autonomous vessel is only as good as the link that binds it to its commander. In a contested environment against a near-peer adversary, the first thing to go isn't the hull; it’s the spectrum.
When the GPS goes dark and the satellite links are jammed, these "autonomous" wonders become very expensive, very quiet drift logs.
We are obsessed with "attritable" systems—the idea that it’s okay to lose a drone because it’s cheap. This creates a dangerous psychological trap. If a commander knows their assets are expendable, they take risks that lead to tactical sloppiness. More importantly, it signals to an adversary that our presence in the water is temporary and fragile.
The Maintenance Mirage
- The Claim: Drones reduce the maintenance burden on the fleet.
- The Reality: Drones shift the maintenance burden to specialized civilian contractors who don't deploy to combat zones.
- The Risk: If a USV breaks down in the middle of the Timor Sea, you can't send a stoker down to the engine room with a wrench. You lose the asset or you risk a manned ship to go retrieve it.
The Sovereign Capability Delusion
The RAN loves the word "sovereign." It implies that Australia is building its own destiny.
The truth? Most of the "brains" inside these autonomous systems are exported from the Silicon Valley ecosystem or filtered through US-based defense giants. By leaning so heavily into autonomous systems before established domestic manufacturing is ready, Australia is merely trading one form of dependency for another.
Instead of being dependent on foreign shipyards, we are becoming dependent on foreign software updates.
Imagine a scenario where a software bug or a "kill switch" in a proprietary algorithm renders 40% of Division 9’s fleet unresponsive during a regional standoff. We wouldn't even have the technical debt to fix it ourselves. We’d be waiting for a patch from a developer in California who is currently on a "mental health day."
Why Decentralization is a Trap
The prevailing wisdom says we should distribute our sensors across hundreds of small drones. "Distributed Lethality" is the buzzword of the decade.
But decentralization increases the attack surface. Every autonomous buoy, every sub-surface glider, and every USV is a node that can be spoofed, captured, or reverse-engineered.
When the Chinese maritime militia "accidentally" snags a RAN drone in a fishing net, what is the protocol? Do we start a war over a piece of plastic and some circuit boards? Probably not. We let them take it. And in doing so, we hand over our sensor data, our encryption standards, and our pride.
By building a fleet of "expendable" drones, we are essentially providing the enemy with a free library of Australian defense technology.
The Recruitment Crisis is the Real Engine
Let’s be brutally honest: The RAN is short thousands of personnel.
The "autonomous" push is a PR shield. It allows the government to say, "We are modernizing," instead of saying, "We can't find enough young Australians willing to spend six months in a metal tube under the ocean."
If we actually had the manning levels required for the Hunter-class frigates and the AUKUS submarines, Division 9 would be a side project. Instead, it’s being positioned as the vanguard.
This is a dangerous gamble. We are betting that AI and remote sensing can replace the intuition of a seasoned Petty Officer. Ask anyone who has actually worked in electronic warfare or sonar: the "eye of the master" cannot be coded. An algorithm can identify a frequency, but it cannot sense the intent behind a maneuver.
Stop Buying Toys and Start Building Systems
The RAN needs to stop treating autonomous units like a tech startup.
If Division 9 is going to be anything other than a taxpayer-funded science fair, it needs to stop focusing on the vessels and start focusing on the data architecture.
Currently, our "autonomous" assets don't talk to each other effectively. A drone from Company A cannot seamlessly hand off a target to a drone from Company B, which in turn cannot update the firing solution on a manned destroyer without three different layers of human intervention and data translation.
The "nuance" the cheerleaders miss is that autonomy without interoperability is just a collection of remote-controlled boats.
The Hard Truths of Autonomous Warfare
- Energy is the Bottleneck: A drone that can stay out for months is a drone that isn't moving fast or doing much processing. High-speed, high-bandwidth autonomy requires power levels that current battery and solar tech cannot sustain in a combat profile.
- Legal Grey Zones: Who is responsible when an autonomous vessel collides with a civilian fishing boat? Or worse, when an automated sensor misidentifies a target and triggers a kinetic response? The RAN has no clear framework for the "algorithmic chain of command."
- The "Silent" Problem: Sub-surface autonomy is a pipe dream in its current state. The ocean is opaque to radio waves. Unless you want your underwater drone trailing a buoy (which makes it a sitting duck), it is operating in a vacuum.
The Path Forward (The One They Won't Take)
If we want a real autonomous edge, we need to stop buying "platforms" and start buying "capability."
That means forcing every vendor to open-source their API so the Navy owns the code, not just the hull. It means integrating these systems into the basic training of every sailor, not just a "specialist unit" tucked away in a corner of the budget.
Most importantly, it means admitting that Division 9 is a stopgap for a failing human pipeline.
The Royal Australian Navy is currently a fleet of grand ambitions held together by shrinking crews and aging steel. Adding a "Ghost Fleet" might look good on a PowerPoint slide, but ghosts don't hold the line. They just haunt the ruins of what used to be a credible force.
The hype is high. The reality is hollow. Division 9 isn't the future of naval warfare; it's the funeral of naval tradition, disguised as progress.
