Cameras on every deck: where Australian fisheries stand as electronic monitoring goes mandatory

A few weeks ago we wrote about the six-week countdown to AFMA's electronic monitoring deadline and what it would mean on the deck. That deadline has now arrived. Between 1 July 2026 and 1 April 2027, electronic monitoring becomes mandatory across the Great Australian Bight Trawl sector, the Commonwealth Trawl sector, the North West Slope and Western Deepwater Trawl fisheries, and the Northern Prawn Fishery. For a large share of the Commonwealth fleet, cameras on the vessel are no longer optional.

It is worth being clear about what this is and what it is not. Electronic monitoring is not a single piece of clever AI. It is a system. Cameras, GPS, sensors on the gear, and storage that records what happens on a vessel during a trip. The footage and data come back to shore, and someone, or something, reviews them to verify what was caught, what was discarded, and what protected species the vessel interacted with. The point is independent verification of the logbook. For decades that verification came from human observers placed on boats, which is expensive, intrusive, and impossible to do at the coverage levels regulators now want. Cameras solve the coverage problem. They create a different one.

The footage problem, and where AI actually helps

The honest difficulty with electronic monitoring has never been the cameras. It is the review. A single vessel on a multi-day trip generates an enormous volume of footage, and most of it shows nothing of interest. Historically, trained reviewers have watched that footage at speed, logging catch events and protected species interactions by hand. As coverage scales up to entire fisheries, the review workload scales with it, and there are not enough trained reviewers to absorb it.

This is the specific gap AI is filling. CSIRO has been developing a computer vision system, WANDA, built to watch fisheries footage and identify catch automatically. The system is trained to recognise species as they come over the rail, count them, and flag the moments a human reviewer actually needs to look at. AFMA has reported that the combination of a near-complete CSIRO algorithm and new review software is already showing the potential to cut footage review time in the Eastern Tuna and Billfish Fishery by a significant proportion. That is the useful, unglamorous core of the story. The AI is not replacing the reviewer. It is doing the first pass so the reviewer spends their time on the events that matter rather than the hours of empty water in between.

For an operator, the practical implication is straightforward. The footage is being watched whether you think about it or not, and increasingly the first viewer is a model rather than a person. Clean, well-lit camera positions and gear that does not obscure the catch are now part of running the boat well, in the same way a legible logbook always was.

Stock assessment: better data in, better decisions out

The value of all this verification is not really about catching anyone out. It is about the numbers that drive how much can be caught at all. Stock assessments, the models that estimate how many fish are in a population and set sustainable catch limits, are only as good as the catch and effort data going into them. When logbook data is independently verified rather than self-reported, the assessments rest on firmer ground.

This is showing up in real work. The Australian Antarctic Division, AFMA, and CSIRO have been jointly reviewing and revising the Patagonian toothfish stock assessment, and that revised work was well received at recent meetings of the Commission for the Conservation of Antarctic Marine Living Resources. Better data does not automatically mean higher quotas. Sometimes it means the opposite. What it means reliably is that the number, whatever it is, is more defensible. For an industry where access to markets and the social licence to keep fishing depend on being able to demonstrate sustainability, a defensible number is worth a great deal.

Enforcement at the edges

There is a third strand worth naming, because it sits slightly apart from the commercial fleet. In the Torres Strait, AFMA has deployed covert surveillance technology to detect and combat illegal fishing activity. This is a different application of the same broad capability: sensors and analysis used to monitor a remote, hard-to-patrol area where illegal activity damages both the resource and the livelihoods of legitimate operators. It is a reminder that the monitoring story is not only about regulating the people who already do the right thing. It is also about detecting the people who do not, in places where putting a patrol vessel is neither cheap nor fast.

The honest assessment

So where does this leave an Australian fishing business in mid-2026? The mandatory electronic monitoring rollout is real, it is happening on a published timetable, and it is not something to wait out. The AI that reads the footage is genuinely working, in the narrow and valuable sense that it reduces the review burden enough to make fleet-wide coverage practical. The verified data is making stock assessments more credible, which is good for the long-term health of the fisheries and the businesses that depend on them.

What this is not is a leap into autonomous, hands-off management. The models still need human reviewers checking their work, the systems still depend on cameras being positioned and maintained properly, and the regulatory framework around how footage is stored, accessed, and used is still settling. There are real questions for operators about data ownership and privacy that have not been fully resolved, and it is reasonable to ask them.

The practical posture is the one we keep coming back to. Treat the cameras as part of the boat, not an imposition on it. Keep the records that make your catch story easy to verify, because increasingly the first thing reviewing that story is a model trained to spot inconsistency. And understand that the verified data flowing off your vessel is feeding the assessments that set your quota. The fisheries that come through this transition well will be the ones that stop treating monitoring as surveillance and start treating it as evidence. The evidence is what keeps the fishery open.

If you are working through what the electronic monitoring rollout means for your operation, or you want a plain account of what the technology can and cannot do, that is exactly the kind of conversation we are having with regional operators now. Get in touch.