Western Australia’s transport authorities have issued more than 30,000 traffic fines since the state‑wide launch of AI‑driven safety cameras earlier this year, prompting the transport minister to plead with drivers to “use common sense” and avoid automated penalties.
AI‑powered cameras roll‑out across the state
The Department of Transport began installing intelligent enforcement units in March 2025 on major highways and arterial roads in Perth, the south‑west regional corridor and a number of country towns. The devices combine high‑resolution video with machine‑learning algorithms that can detect a range of offences, including speeding, red‑light running, non‑use of seat‑belts, mobile‑phone use and tail‑gating. Once an infraction is recorded, the system automatically registers the vehicle’s registration details, generates a citation and issues a fine through the same online portal used for conventional traffic tickets.
According to a Reuters report, the technology is built on the same AI models used in other safety‑critical applications, such as autonomous‑vehicle perception systems, and has been calibrated to meet Australian privacy standards.
Transport Minister’s public appeal
Transport Minister Rita Saffioti addressed the media on Tuesday, urging motorists to “pay attention to the road, obey the speed limits and keep distractions to a minimum.” She warned that the AI system “does not forgive careless behaviour” and that “common sense remains the cheapest way to avoid a fine.” Saffioti’s comments were part of a wider communications push that included television spots and digital advertisements featuring the slogan “Drive smart, avoid the AI ticket.”
“We are giving drivers a clear, transparent system that works 24/7,” Saffioti said. “If you respect the rules, the AI won’t have a reason to penalise you.”
Technology behind the enforcement
Each camera unit houses a suite of sensors: radar for speed measurement, infrared for night‑time detection, and a pair of stereo lenses that feed visual data into the AI engine. The algorithm has been trained on millions of footage frames collected over a two‑year trial period, allowing it to differentiate between genuine infractions and false positives such as emergency vehicles or road‑work diversions.
When an offence is flagged, a human operator reviews the flagged footage for verification before a citation is dispatched. The Department of Transport says this “human‑in‑the‑loop” step reduces the risk of erroneous fines, a concern raised by privacy advocates during the system’s consultation phase.
Impact on drivers and revenue
Since the system became operational, the state has recorded 30,487 fines, generating approximately A$13 million in revenue. The most common violations are speeding (57 %), followed by failure to wear a seat‑belt (22 %) and illegal mobile‑phone use (14 %).
Data released by the Department show a 12 % reduction in average speeds on the monitored corridors compared with the same period in 2024, supporting the government’s claim that the technology improves road safety.
Expert and community response
Road‑safety researcher Dr. Amelia Patel of the University of Western Australia welcomed the early results, noting that “any reduction in speed translates into fewer severe crashes, and AI enforcement provides a consistent deterrent that traditional policing cannot match.”
Conversely, the Australian Digital Rights organisation has warned that the system could set a precedent for broader surveillance. “We must ensure that data retention policies are strict and that drivers have a clear avenue to contest a ticket,” said spokesperson Jordan Lee.
Legal and regulatory considerations
The introduction of AI‑based traffic enforcement arrives amid a global debate over algorithmic accountability. In the United Kingdom, the Competition and Markets Authority is preparing to force Google to disclose how its AI models affect competition, a move that underscores growing scrutiny of AI’s societal impact (see Reuters). Australian regulators have indicated they will monitor the system for bias or disproportionate impacts on certain driver groups.
Motorists can appeal a fine within 28 days through the Department’s online portal, providing evidence such as GPS logs or dash‑camera footage. The appeal process is overseen by an independent tribunal, which, as of the latest quarter, has upheld 94 % of the AI‑issued tickets.
Looking ahead
The state plans to extend the AI‑camera network to an additional 150 sites by the end of 2026, focusing on high‑risk intersections and rural highways. Transport Minister Saffioti has also announced a public‑education campaign that will include free safety workshops in community centers.
Meanwhile, the Department is evaluating next‑generation capabilities, such as real‑time traffic‑flow optimisation and integration with the upcoming “Smart Roads” initiative, which would allow dynamic speed‑limit adjustments based on live congestion data.
For an in‑depth look at how AI regulation is shaping Australian policy, see our continued coverage on AI regulation in Australia.
