For more than a century, the internal combustion engine was the beating heart of the automotive world, a mechanical marvel honed over generations. But under the bonnet of the modern car, something fundamentally different is stirring. It is not just horsepower that is being re-engineered; it is brainpower. Artificial intelligence is no longer a futuristic gimmick reserved for concept cars. It has become the industry’s new assembly line, its chief designer, and its most controversial foreman.
From the high-tech design studios in Seoul and Detroit to the dealership floors of Bryanston, AI is reshaping what a car is and how it is made. In South Africa, this shift is already tangible. Hyundai Automotive South Africa recently revealed that its proprietary AI platform, HANA, has processed over 70,000 online chats in a single year, proving that even in a relationship-driven market, the machine has become the first point of contact.
But as the industry races towards fully autonomous vehicles and lights-out factories, a critical question lingers: In the race to optimise every line of code and every millimetre of design, where does the human fit in?
Looking for a safe car for a student then click here
Breaking the Pencil
For decades, the automotive design process was a ritual of clay, drafting pencils and subjective human taste. That world is vanishing faster than a lead-footed V8 at a petrol station. AI has moved from a supportive tool to a creative co-pilot.
Ford’s Chief of AI, Franziska Bell, recently described the shift as turning “manual hours into pushes of a button.” Engineers can now feed constraints—like weight, material cost and aerodynamic drag—into generative algorithms. The machine then spits out hundreds of design variations that no human hand could have sketched.
“We can generate hundreds of variations at a push of a button,” Bell explained, noting this elevates designers to “executive directors” of the process rather than grunt workers.
This is more than just styling. It is structural alchemy. Researchers recently showed a generative design process on a go-kart steering wheel, using AI algorithms to cut the component’s weight by a staggering 60% while keeping structural integrity. The computer found organic, almost skeletal geometries that a human engineer might never have considered.
However, the rise of the ‘cyborg designer’ comes with a hangover. There is a growing fear of ‘creative-skill erosion’. If a machine generates the baseline concept, who owns the intellectual property? And if an AI optimises a part for cost but misses a rare edge case in safety, who is held liable?
Starting or running a small business and in need of a bakkie – click here
Efficiency vs. Displacement
The hum of the smart factory is quieter, but infinitely more data-dense. Predictive maintenance algorithms analyse the vibration of robotic arms to stop breakdowns before they happen. Computer vision systems inspect welds with a precision the human eye cannot match. McKinsey estimates these smart systems yield efficiency gains of 20%–30%.
But this efficiency comes with a human cost.
In May 2026, General Motors delivered a stark warning to the workforce. The automaker cut 600 IT workers while simultaneously launching a hiring spree for AI specialists, offering salaries that dwarf traditional automotive wages. It is a brutal ‘blood transfusion’ for the industry; the technical skills of yesterday are being traded for the algorithmic literacy of tomorrow.
Great insurance is a safety net you really need – click here to find out more
Johan Nel, Sales Director at Hyundai Automotive South Africa, is acutely aware of this tension. Speaking at a conference in Bryanston, he insisted the company’s strategy is human-centric.
“At Hyundai Automotive South Africa, we believe innovation should empower people, not replace them,” Nel said. However, the company’s own data illustrates the shifting landscape. Since deploying HANA, digital dealer sales have increased by 280%. While this drives revenue, it also fundamentally changes the job description of the salesperson, moving them from transaction-closers to experience-managers.
The ‘Jerk’ Factor
You are just one click away from finding the very best finance deal for your new car
Inside the car, AI is trying to be less of a computer and more of a companion. From personalised seat settings to voice assistants like Stellantis’ ‘Cara’, the vehicle is learning the driver’s habits. But the integration of AI into the Human-Machine Interface (HMI) is fraught with peril.
The industry calls it the ‘human-automation interaction failure’. What happens when the driver trusts the AI too much? Or not enough?
Research into adaptive HMI—originally developed for military fighter jets—is now trickling down to passenger vehicles. The goal is ‘adjustable autonomy’, where the AI handles the boring highway cruise but immediately surrenders control to the human when the road gets twisty or the weather turns foul.
The challenge is validation. How do you prove an AI is safe for the ‘edge case’—the drunk pedestrian jumping out from behind a taxi, or the pothole hidden by a puddle? Engineers are increasingly relying on simulation, running millions of virtual miles to stress-test the software. But as the SAE International notes, while simulation accelerates validation, it cannot fully replace the unpredictability of the real world.
The Road Ahead
The lack of a unified regulatory framework is perhaps the industry’s greatest headache. In the US, oversight is fragmented. In the EU, the AI Act imposes strict risk-based standards. In South Africa, the industry is largely self-regulated, following global standards set by manufacturers.
The bottom line is this: AI is no longer just a feature on a spec sheet. It is the organisational principle of the entire auto industry.
The winners in this new era will not necessarily be the companies with the fastest chips or the biggest datasets. They will be the firms that solve the ‘governance equation’—those that pair aggressive technical adoption with disciplined ethics, transparent safety validation, and, as Hyundai is attempting in South Africa, credible plans to reskill their workforce for a world where the line between the driver and the driven is forever blurred.
As Stanley Anderson, CEO of Hyundai Automotive South Africa, admitted regarding the HANA rollout: "It was not always smooth sailing... we encountered initial fluctuations." But they persevered, optimised, and scaled.
In the race to automate, the ability to pause, extract insight, and put the human back in the driver’s seat might just be the most valuable intelligence of all.
Colin Windell for Colin-on-Cars in association with
proudly ALL THINGS MOTORING
