Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
When you are driving in your car, are you in a public or private space? As carmakers race to make electric cars into smartphones on wheels, the lines have started blurring.
Internet-connected intelligent cars are on the roads in many countries, with remote car access an increasingly common feature. In some cities in China, driverless taxis are already available. But with convenience follows concerns over who has access to our driving data — and ultimately our cars.
The Biden administration sees Chinese cars and technology as a national security threat that could spy on drivers and as an economic threat, and has moved to ban Chinese connected-car software. Beijing has previously accused the US of using “national security” concerns as an excuse to act against Chinese companies. Indeed, the economic implications of such a ban on software are clear: it would in effect keep Chinese EVs out of the US. Current tariffs on Chinese EVs are easy enough to circumvent by shifting production bases. But given the scale of personal data that is expected to be gathered from connected cars — not just those made in China, but by automakers globally — the latest proposed ban should not be brushed off as just another trade spat. It raises important questions about privacy and safety for drivers around the world.
As EV sales grow, cars are increasingly becoming software-defined vehicles — that is, any car that uses software to operate and add upgraded functionality. Around 97 per cent of all EVs are internet-connected.
Connected intelligent cars offer many benefits. Safety features such as anti-collision systems, real-time data analysis and advanced sensors mean faster reaction times and fewer accidents. Autonomous driving functions help provide mobility for elderly drivers. Carmakers upgrade cars using over-the-air updates. Connected cars using better routes can mean fewer traffic jams, reducing emissions.
Chinese EV makers are leading the race. Development times for new models have been about 30 per cent quicker than legacy peers. More than 60 per cent of new energy vehicles sold in China this year had advanced driving-assist features, according to industry data.
Take Apollo Go, the robotaxi arm of Chinese internet group Baidu. Its cars can handle most driving situations independently without a driver. More importantly, Baidu has brought costs down significantly, with its latest sixth-generation robotaxi costing less than half its previous model at about $28,500.
Getting to this point, however, has been costly. Baidu started investing billions from its autonomous driving fund in 2017. Its self-driving project has required years of testing and is still lossmaking. Chinese EV maker BYD is set to invest $14bn in areas related to autonomous driving to catch up. Local peer Nio, a leader in autonomous driving software, also remains lossmaking despite growing sales.
That means some global automakers in a rush to catch up on intelligent driving software, or lack the funds to develop their own, may start to turn to software that includes Chinese technology. Without that option, carmakers risk falling behind on innovation and a lucrative market — the market for robotaxis alone is expected to exceed $25bn globally by 2030, according to Goldman Sachs. By 2027, it forecasts partially autonomous cars, which require driver supervision, are expected to be about 30 per cent of global new car sales. For buyers, affordable EVs may become further from reach as development costs rise.
As cars are such an integral part of life for many of us, the potential risks are amplified — perhaps more so than with any other product. For carmakers to provide remote assistance and upgrades after the sale, and for self-driving functions using cameras and sensors, cars must be connected in real-time. Improving software also requires extensive data collection. That means that there will be a risk — however small — that a connected car could be affected by a cyber attack or data breach. Functions could be accessed remotely, affecting driver safety. Sensors can be manipulated to detect false objects on the road. A hacked self-driving car could even be turned into a weapon.
The race towards smarter, self-driving cars is starting to outpace discussions on privacy, cyber security risks and regulation. Closing trade borders might mean depriving some countries of the latest innovations. But until governments find the balance of risk and reward it may be worth taking the slow road.