By 2027, autonomous vehicles (AVs) will redefine workplace safety in Australia—not as a technological novelty, but as a legal minefield where employers face non

•By 2027, autonomous vehicles (AVs) will redefine workplace safety in Australia—not as a technological novelty, but as a legal minefield where employers face non
By 2027, autonomous vehicles (AVs) will redefine workplace safety in Australia—not as a technological novelty, but as a legal minefield where employers face non-delegable liability. This isn’t just regulatory evolution; it’s a strategic realignment where businesses must now treat AVs as core infrastructure, not just tools. The stakes? Legal accountability that could eclipse even the liability battles of the early automobile era. Here’s what I find interesting: the same companies pioneering automation are now racing to build safety frameworks before regulators force them to.
Australia’s Bureau of Infrastructure and Transport Research Economics (BITRE) projects 2.6% of new vehicles will be fully automated by 2030, surging to 50% by 2046. But the critical inflection point comes sooner: full deployment is anticipated by 2027. This timeline creates a regulatory race. The proposed AV Safety Law (AVSL) framework introduces Automated Driving System Entities (ADSE) liability, but employers remain locked in a paradox. While manufacturers and software providers share legal exposure, the WHS Act’s non-delegable duty means employers cannot outsource their primary responsibility under Section 19. As BITRE’s data shows, the clock is ticking—businesses in logistics, mining, and construction have just five years to adapt.
“The AVSL isn’t just a law—it’s a liability reset button for industries relying on autonomous systems.”
Yet the gap between regulation and reality is stark. The law’s “reasonably practicable” standard remains vague, leaving employers to guess what courts will demand. This uncertainty is a ticking time bomb for industries already grappling with legacy safety systems.
Employers face a stark reality: they cannot delegate safety obligations even when sharing liability with manufacturers. Under the WHS Act, Section 16 clarifies that concurrent liability doesn’t absolve duty holders. This creates a strategic dilemma. For example, a mining company deploying autonomous haul trucks must still conduct failure mode assessments and cybersecurity audits, even if the AV’s software is developed by a third party. The legal chessboard here is clear: employers must now control what they can influence, or risk bearing full liability.
Dentons’ analysis underscores the challenge: courts will interpret “reasonably practicable” through the lens of emerging industry norms—a moving target requiring constant vigilance. I’ve been tracking this space for years, and this is the first time I’ve seen employers forced to treat software supply chains as safety-critical infrastructure. The old playbook of outsourcing responsibility is dead.
Traditional safety systems focused on training and supervision. AVs demand a paradigm shift. Employers must now implement “safe systems of work” that include:
In my view, this signals a shift from reactive compliance to proactive engineering. Companies like NVIDIA and DeepMind are already embedding safety into AV algorithms—a model Australian businesses must adopt. The Department’s proposed reforms hint at this future: safe systems will be judged not just on accident prevention, but on how well they anticipate edge cases. Imagine a self-driving truck navigating a sudden debris field in the Pilbara mines—this is where liability will be tested.
Waiting for legislative clarity is a losing strategy. The WHS Act’s existing “plant” definition already classifies AVs as employer-controlled assets. Delaying action risks:
Smart employers are already moving. Logistics firms are partnering with cybersecurity experts to audit AV software supply chains. Mining companies are stress-testing autonomous systems in virtual environments—a practice pioneered by Microsoft’s Azure Digital Twins. The next 18 months will see a surge in third-party compliance audits, as businesses preemptively build defensible safety frameworks.
Watch Apple Intelligence and Meta’s Llama teams in 2025. Their AI-driven safety modeling could set new industry benchmarks, forcing laggards to play catch-up. This isn’t just compliance—it’s a strategic arms race. The question isn’t whether AVs will transform Australian workplaces, but who will control the rules of engagement.
— Romaric Anderson, Tech Curator at AI Loop
Your feedback directly trains our AI agents to improve.