A serious shark bite at Coogee has turned attention to whether AI-assisted drones could help detect dangerous sharks faster and give swimmers earlier warnings.
Coogee Attack Renews Focus On Shark Warnings
Coogee Beach has become the centre of renewed attention on shark surveillance after a serious attack left a swimmer critically injured and brought drone monitoring back into public view.
The swimmer, Leah Stewart, a 35-year-old mother and teacher, was attacked by a great white shark while swimming between the flags at Coogee on Saturday, 13 June. She has undergone multiple surgeries and had her left arm amputated.
The beach was closed after the incident and later reopened with drone surveillance. The reopening has placed fresh focus on how drones, artificial intelligence and public alerts could work together at popular Sydney beaches.
Scientists have said an early warning system using autonomous drones, onboard sensors and AI shark recognition could be possible by summer if existing shark safety programs were expanded.
AI Drones Could Strengthen Coogee Surveillance
Macquarie University Professor Culum Brown has said autonomous drones fitted with onboard sensors and artificial intelligence could detect sharks from the air.
In a more advanced version, a drone could identify a shark, track its movement and send information back to a base. That information could then help trigger warnings for swimmers to leave the water.
Emeritus Professor Rob Harcourt, also from Macquarie University, has said autonomous drones and AI shark recognition technology are either available or close to available.
The technology would not replace lifesavers or drone pilots. It would add another layer to existing patrols and help beach teams respond more quickly when dangerous sharks are seen close to shore.
Flight Restrictions Kept Drones Away From Coogee
Coogee did not have drone use before the attack because of restrictions linked to the area being under a flight path. An emergency exemption was granted after the incident, allowing drone surveillance to operate at the beach.
The situation has drawn attention to uneven drone coverage across Sydney’s coastline, where patrols depend on location, permissions, season and conditions.
Drone patrols across NSW reach up to 80 sites during the summer peak before dropping back during winter. A separate extension of daily drone surveillance from 27 April to 30 June 2026 covered 14 locations, including Maroubra Beach, but Coogee was not listed.
Shark sightings and beach closures from drone patrols are uploaded to the SharkSmart app, giving beachgoers updated information before entering the water.
Belrose Operations Sit Within The Wider System
Surf Life Saving NSW chief executive Steven Pearce has said shark identification is still handled by human operators, either on the beach or from head office, while work continues on finding an AI system with acceptable accuracy.
Surf Life Saving NSW also runs some autonomous drone operations from its head office in Belrose for remote areas, showing how drone surveillance can be supported both on the beach and away from the shoreline.
The wider shark safety network already includes drone patrols, SMART drumlines, tagged shark listening stations, shark nets, the SharkSmart app, education campaigns and community shark bite kits.
Accuracy Will Shape Any AI Rollout
The future of AI-assisted shark detection will depend heavily on accuracy.
Too many false alarms could lead to repeated beach evacuations if dolphins, grey nurse sharks or wobbegongs are mistaken for dangerous sharks. Beach safety teams have been working towards an AI solution that can identify sharks with acceptable reliability.
Coogee’s reopening under drone surveillance has made the technology more visible along the beach. The longer-term question is whether AI detection can help close the gap between a shark being spotted and swimmers being warned.
The issue now sits at the shoreline: how to detect sharks sooner, how to warn swimmers clearly and how to use new technology without overstating what it can guarantee.
Published 16-June-2026
