Acusensus is an ASX listed AI surveillance company with a five year, $100 million contract to operate New Zealand's mobile camera network. Its own investor presentations describe a business model built on expanding the scope of what it monitors. New Zealanders are already pushing back.
In February 2026, someone drove up to a speed camera on Coatesville Riverhead Highway north of Auckland, stopped their blue 4WD, sliced the pole clean through with an angle grinder, and left before police arrived. Rodney local board member Geoff Upson filmed the aftermath and posted it to Facebook. Standing beside the toppled pole, he called it "the revenue collector" and told his followers it was safe to do 61km per hour on the way to work. 1News reported the camera had collected more than $10 million in fines since the speed limit on that stretch dropped from 80km to 60km in June 2020.
Five cameras were damaged or destroyed across Auckland in the same week. NZTA called it criminal damage and reported the incidents to police. Most New Zealanders watching the footage called it something else.
This is where we are. And to understand how we got here, you need to understand exactly who Acusensus is, how they make their money, what their own documents say about where this is heading, and what every driver in New Zealand should know about the company now operating the country's mobile camera fleet.
Who Is Acusensus?
Acusensus is a Melbourne based technology company founded in 2018 by Alexander Jannink, a former head of the Future Product Group at Redflex Traffic Systems, and Ravin Mirchandani, the CEO of Indian industrial conglomerate Ador Powertron, which provided the seed funding and whose founder now chairs the company. Acusensus listed on the ASX in January 2023 at a market capitalisation of approximately $100 million Australian dollars. Its ticker is ACE.
Jannink has spoken publicly about his founding motivation: a close friend was killed by a distracted driver. That personal impetus is genuine and it drove the company's initial focus on mobile phone detection cameras, which it pioneered in New South Wales from July 2021 in what the company describes as the world's first live enforcement program for distracted driving.
The business model is not revenue per ticket. Acusensus collects a fixed monthly fee from government clients under multi year contracts. As Jannink told investors at the ASX SMIDcaps Conference in March 2025, "we supply these automated enforcement cameras and we receive a fixed monthly fee essentially for those services." NZTA has confirmed this structure applies in New Zealand: "Neither Acusensus or NZTA will receive any incentives or funds from tickets issued."
The New Zealand contract is worth approximately $100 million over five years, or around $20 million per year, confirmed by RNZ's investigation published in March 2025. For context, Acusensus recorded trailing twelve month revenue of $38.4 million Australian dollars as of June 2025, per PitchBook financial data. The New Zealand contract alone represents a dominant portion of the company's revenue base.
That is the foundational fact most New Zealanders have not been told. This is not a government agency operating public infrastructure. This is a publicly listed private company, accountable to shareholders, deploying artificial intelligence surveillance across New Zealand roads on a $100 million government contract, with its entire business model built on winning and expanding those contracts.
The Technology: What the Cameras Actually Do
The Acusensus Harmony system deployed in NZTA's fleet of unmarked SUVs and trailers uses continuous radar transmission detecting vehicles from up to 300 metres with a typical operating range of 150 metres, paired with high definition still image cameras. This has been confirmed via Official Information Act request and published by 1News.
For speed enforcement purposes in New Zealand, the cameras photograph speeding vehicles. The system does not currently use automatic number plate recognition in its deployed configuration here. NZTA makes the final decision on whether an infringement is issued.
But the Harmony system is capable of considerably more than this.
In Australian deployments where Acusensus operates phone detection programs, the cameras photograph every passing vehicle. As RNZ describes: a machine learning AI then combs through each image looking for signs of a driver not keeping both hands on the wheel. If that condition is triggered, the system looks for an object matching the dimensions and appearance of a mobile phone. Images with no such signs are deleted. The ones that pass are packaged as encrypted evidence, uploaded to Acusensus servers on the Amazon Web Services Cloud database in Australia, and sent to the government client as confirmed evidence of an offence.
That is the full capability of the cameras now operating on New Zealand roads. It is not what they are currently doing here. But it is what they can do.
As RNZ confirmed, NZTA ran trials using this exact Acusensus phone and seatbelt detection technology on Auckland motorways in 2022. The official position now is that current New Zealand legislation does not permit cameras to be used for anything beyond speed and red light enforcement. NZTA told RNZ: "NZTA has not discussed this use of safety cameras with Acusensus." What they did not address is that the Acusensus 2022 trial for phone and seatbelt detection in New Zealand already happened. The legal prohibition on expanding the scope can be changed. Laws change. Contracts expand.
Like this? Get more MotorBuzz delivered to your browser. Click here to add MotorBuzz as a preferred news source on Google.
The Investor Presentation Nobody Is Talking About
In its 2024 investor presentation, Acusensus stated its business model in terms no New Zealand government press release would use. The company told its shareholders:
"Over time, contracts are awarded, and then those contracts are typically expanded."
That is not spin. That is a statement to investors about how the company grows revenue. You start with speed cameras. You expand to phone detection. You expand to seatbelts. You expand to tailgating. You expand to impaired driving. Each expansion requires a contract variation or a new contract. Each generates more monthly fee revenue.
Jannink was equally direct at an Australian investor event in 2023. As quoted by RNZ: "There's a lot of artificial intelligence development to do all these new enforcement activities of, you know, mobile phone, seatbelt, and we are applying the smarts of the team to trying to solve other challenges like impaired driving. How do we automatically work out who on the road network is impaired by drugs or alcohol?"
This is a company that, in its own words, is developing AI to identify potentially impaired drivers from camera images. In Queensland it is already testing cameras that detect tailgating. In Western Australia it is trialling detection of variable speed limit breaches. In the UK, what police there described as "a world first" test of AI camera technology targeting drunk drivers began in December 2025, as reported by RNZ.
The cameras parked outside your child's school are the entry level product.
The Privatisation of Policing
Before July 1, 2025, New Zealand Police operated mobile camera vans. Officers, subject to the Police Act, their code of conduct, and direct public accountability, decided where cameras went and when they operated. Complaints about Police conduct have a formal resolution pathway. Officers are individually accountable.
From July 1, that responsibility transferred entirely to NZTA, which contracted operations to Acusensus NZ. As confirmed in NZTA's OIA response documents, fixed camera operations are run by Redflex, trading as Verra Mobility. Mobile operations are run by Acusensus NZ. The human beings operating equipment that issues $150 fines to New Zealanders work for a Melbourne based ASX listed company.
This is not a minor administrative change. It is the transfer of a coercive public function, the power to fine citizens on public roads, to a private entity whose primary obligation is to its shareholders. NZTA insists the decision to issue a fine remains with the agency, not with Acusensus. That is technically true and practically incomplete: the evidence, the detection, the image capture, the initial review and the chain of custody are all owned and managed by a private company with a financial interest in demonstrating the value and effectiveness of its technology to secure future contracts and contract expansions.
The screenshot circulating on New Zealand social media from Acusensus's own website states that 99% of captured images are deleted and never seen by a human, that image reviewers at the first stage of review cannot identify the location, time, day, registration, or other vehicle occupants, and that full original evidence is typically only accessed if images go to court. Acusensus presents these as privacy protections. A different framing is equally accurate: an AI system is photographing every vehicle that passes within 300 metres of an unmarked SUV, making autonomous judgements about each image, and the data sits on servers in Australia on an Amazon Web Services cloud database.
What Other Countries Did When Drivers Pushed Back
New Zealand's public anger is not unique. The global pattern is consistent enough that it constitutes evidence of how this ends if governments do not respond.
The first speed camera programs in the United States, installed in Friendswood and La Marque, Texas in 1986 and 1987, lasted only a few months each before, as Wikipedia documents, public pressure forced them to be dropped entirely.
In the UK, Department for Transport guidance requires cameras to be visible, signed, and justified by crash data at the specific location. As UK road law barristers have noted, the guidance itself carries no legal force but has historically been the political floor below which camera programs cannot fall without triggering significant public opposition. The UK's yellow camera convention exists precisely because of public pressure following the early proliferation of grey and unmarked devices.
France introduced its automated enforcement network in 2003 following a road safety crisis, but the placement framework, documented by the European Commission road safety research programme, requires cameras to be sited on accident prone roads with published crash data justification. As Ontario Premier Doug Ford demonstrated when he announced legislation in October 2025 to ban speed cameras altogether after the Toronto vandalism wave, calling them "cash grabs," the political risk of unmarked, undisclosed, AI operated cameras without crash site justification is severe.
The Virginia Senate advanced bills in January 2026 on 8 to 7 votes, with Democrats crossing the aisle, to eliminate speed and red light cameras statewide. Richmond saw four of eleven school zone cameras spray painted in a single week in February 2025. Des Moines installed security cameras on its speed camera trailers specifically to stop them being destroyed. These data points, all reported and tracked by MotorBuzz, trace an identical arc: stealth deployment, revenue generation, public fury, vandalism, political capitulation or escalation.
The research on what actually reduces deaths rather than generates revenue is not ambiguous. A meta analysis from the College of Policing, covering 51 studies, found no evidence that the effect of speed cameras on speeding behaviour or collisions differed by whether cameras were covert or overt. The safety case for stealth is not supported by the evidence. What stealth achieves is catching more drivers, not changing more behaviour. Catching and changing are different outcomes with different consequences, and only one of them actually saves lives at scale.
The European Commission speed enforcement research is explicit that visible cameras produce deterrence effects that dissipate after three days. Fixed cameras at proven crash sites produce the most durable safety benefit. Mobile cameras, particularly unmarked ones operating outside designated black spots, produce fine revenue and public rage in roughly equal measure.
The Questions NZTA Has Not Answered
The legitimate questions are not about whether speeding kills people. It does. They are structural questions about accountability, transparency, and the direction of travel that no official spokesperson has directly addressed.
Who approved the decision to hand a $100 million contract to a foreign private company with explicit investor communications about contract expansion as the growth model? What safeguards exist against scope expansion beyond speed enforcement beyond the current legislative prohibition, which can be changed by any future government? What is the process for a New Zealand citizen to challenge the accuracy of an AI detection system operated by an Australian company using servers in Australia? If Acusensus expands into phone detection or impaired driving detection in New Zealand, at what point does that require fresh public consultation rather than a contract variation signed by NZTA?
NZTA has confirmed the contract term and the operational rules. It has not published the contract itself. It has not provided a public register of deployment locations or the crash data justifying each site. It has not committed to a review process if detection scope expands.
The company that photographed 99% of cars passing its cameras last week and deleted all but a small fraction owns that data infrastructure. It retains the technical capability to deploy its full detection suite on New Zealand roads the moment the legislative prohibition lifts or is amended. And its own investor communications confirm that contract expansion is the explicit business strategy.
What You Can Do
Speeding is dangerous. Nobody in this article is arguing otherwise, and MotorBuzz has covered the genuine road safety evidence in detail. The argument is about where cameras go, who operates them, whether the placement decisions are genuinely driven by crash data, and whether handing that function to a listed private company with an expansion mandate represents good public policy.
Drivers who receive notices they believe were issued from cameras operating outside NZTA's own published rules, as detailed in our previous guide to disputing NZ speed camera notices, have a legitimate right to dispute. NZTA's operational guidelines create an enforceable standard. A notice from a camera positioned within 250 metres of a passing lane end, on a curved road section, or from a device unable to produce a valid calibration certificate within the preceding twelve months, is a notice worth challenging in writing before the deadline.
Beyond individual tickets, the political pressure is the only mechanism that has historically changed how these programs operate. Ontario, Virginia, Texas, the UK, France and the Netherlands all modified their programs in response to public pressure and political accountability, not in response to arguments about road safety evidence.
In New Zealand, that pressure needs to be directed at the decision to privatise enforcement in the first place, at the absence of published crash data justifying specific deployment sites, and at the unanswered question sitting in every Acusensus investor presentation: what exactly does contract expansion look like when it reaches New Zealand?
The cameras are watching. The question is whether New Zealanders are watching back.