Australia's Seatbelt Moment: How 1970 Road Safety Can Shape AI Defense

In 1970, Victoria recorded 1,061 road deaths and responded by becoming the first jurisdiction in the world to mandate seat belts — and a new commentary argues Australia must now apply that same engineered-safety logic to artificial intelligence.

The piece, framed around the federal parliamentary inquiry into AI, contends that the country's public safety history offers a direct template for governing autonomous software. Victoria's road toll fell by more than 70 per cent over the three decades after 1970, a decline the author attributes to combining human accountability with compulsory, engineered physical controls. Airbags and crumple zones later supplemented the belt. The lesson, the argument runs, is that rules alone never fixed the road toll — protection built into the machine did.

Why does the seatbelt analogy matter for AI?

Australians embraced the automobile through the 1950s and 1960s, but by 1970 faster vehicle speeds had pushed Victoria's annual road toll to 1,061. Public education and traffic rules could not curb the epidemic on their own. Legislators recognised that road safety required combining human accountability with compulsory, engineered physical controls, and in 1970 Victoria made global history as the first jurisdiction to mandate seat belt wearing.

The author draws a straight line from that precedent to today's AI security problem. Regulatory guidelines provide necessary boundaries, the piece states, but they cannot stop an attack that tricks an AI system into following malicious instructions, nor halt a breach already in progress. The required shift is from rules governing how technology is used to protections built into the technology itself.

The stakes have grown because AI systems increasingly act on their own. Software now operates autonomously and with authority across multiple government systems — from breaching public-facing sites to finding vulnerabilities in critical administrative platforms, the commentary notes. This marks a change in the governance question itself.

"The fundamental governance question facing Australian leaders is no longer about controlling what software can see — it is governing what software is authorised to do."

What did the parliamentary inquiry expose?

The Australian Federal Government's parliamentary inquiry into Artificial Intelligence, according to the commentary, laid bare a stark reality: acknowledging the risks of autonomous, rogue AI agents after a breach occurs is not sufficient. The author warns that Australia must capture the full economic and social benefits of AI without opening a dangerous operational security gap.

The prescription is specific. The country should move beyond high-level governance promises and embed "Secure AI by Design" principles into AI systems from the outset, supported by continuous, real-time technical safeguards. The inquiry's recommendations, the piece argues, give Parliament an opportunity to align governance frameworks and technical defences with the speed of the technology itself.

What makes AI supply chains an unseen threat?

Artificial intelligence, the commentary asserts, might be the most significant new security challenge facing Australian organisations — but it also offers the most powerful new tools for their defence. Realising the second half of that equation requires organisations to fundamentally rethink security practices.

Autonomous AI agents introduce threat categories that existing tools were never designed to manage:

• Agents can be tricked into revealing sensitive information.
• Agents can impersonate legitimate users.
• Agents can take actions they were never authorised to perform.

Traditional cybersecurity tools were not built for any of these scenarios, the author states.

Speed compounds the problem. As adversaries use AI to compress cyber-attack timelines from days down to minutes, manual approvals and periodic audits cannot keep pace. The attack window that human review once covered has collapsed to a span no human reviewer can cover.

The construction of modern AI systems adds a further layer of exposure. AI applications are rarely built from scratch. They are assembled from models, data and software components developed by numerous third parties, and each component can introduce its own security risks. The consequence, in the author's words, is that AI models and the tools they use must be checked before deployment, with their behaviour continuously monitored as systems evolve.

How can AI defend against AI?

The commentary's central turn is that the same capabilities powering attacks can power defence. The speed and reasoning capabilities that make AI a potent tool for threat actors can be turned inward to protect digital infrastructure.

Under expert human supervision, advanced AI systems can help cybersecurity teams in three concrete ways: discover weaknesses, understand how attackers might exploit them, and rapidly put protections in place. Centralised security systems can monitor and control how AI applications interact with organisational data and other systems, keeping their actions authorised and accountable.

This defensive speed matters most where downtime is not an option. The piece identifies energy grids, transport networks and healthcare systems as critical infrastructure whose operational systems cannot easily be taken offline for conventional patching. For these sectors, rapid defensive capability is described as a vital lifeline.

What role should the AI Safety Institute play?

The commentary identifies a gap in current policy: defenders cannot mitigate risks they cannot analyse in advance. For Australia to stay ahead of machine-speed threats, government policy must facilitate early, managed access to cutting-edge frontier models for three groups — national defenders, security researchers and critical infrastructure operators — before those models are broadly deployed.

Australia's AI Safety Institute already has an important role in evaluating emerging risks, the author acknowledges. But that role should extend further: the Institute should help ensure that government agencies, security researchers and critical infrastructure operators can safely access advanced AI capabilities for defensive purposes. The argument treats pre-deployment access for defenders as the informational equivalent of crash-testing a car before it reaches showroom floors.

Can Australia set a global standard twice?

The piece closes on the historical parallel it opened with. Just as mandated seatbelts saw Australia lead the world in vehicular safety half a century ago, the country now has the opportunity to set a global standard for secure AI adoption.

The author frames the path as a balance between innovation and safety, achieved through three commitments: adopting continuous technical safeguards, giving defenders timely access to advanced AI capabilities, and ensuring those responsible for protecting critical systems can respond as quickly as attackers.

Done well, the commentary concludes, this approach will allow Australia to realise the benefits of AI while protecting national security and maintaining public trust — a outcome with direct precedent in the road-safety record Victoria built after 1970.

The question the parliamentary inquiry's recommendations now put to government is whether Australia repeats its seatbelt precedent in software: building protection into the machine before the toll, not after it.