Self-driving cars sound amazing, right? You sit back, the car does the driving, and traffic jams become a little less painful. But there’s one question people always ask:
“Can a self-driving car be hacked?”
The short answer is yes, in theory autonomous vehicles can be hacked. But before you picture a hacker steering your car from a dark basement somewhere, let’s slow down. Real-world car hacking is incredibly hard to pull off, and modern vehicles are built with layers of protection that make it even harder. Still, it’s worth understanding what “hacking” actually means in this context, and how manufacturers are keeping drivers safe.
How Could Someone Hack a Self-Driving Car?
Modern cars are a bit like smartphones on wheels, full of sensors, apps, and internet connections. These are what make them smart, but they also create opportunities for hackers if not properly secured.
Here are the main ways it could happen, and what carmakers are doing to stop it.
1. Software Bugs or Weak Points
Self-driving systems rely on millions of lines of code. Every so often, researchers find bugs that could, in theory, be exploited.
Back in 2015, two security researchers made headlines when they managed to remotely control a Jeep through its entertainment system [1]. It was a wake-up call for the industry, and led to new cybersecurity rules and much tighter software checks.
How it’s fixed: Cars now have digital firewalls separating safety systems from infotainment, plus regular security updates to patch any vulnerabilities.
2. Tricking the Car’s Sensors
Self-driving cars “see” the world through cameras, radar, and lasers (called LiDAR). Researchers have shown that these sensors can be fooled. For example, a laser or oddly-shaped sticker on a stop sign can make a car think it’s seeing something else [2].
How it’s fixed: Cars now use sensor fusion, meaning they cross-check what one sensor sees against others. If something doesn’t add up, the system ignores it.
3. Messing With GPS or Communications
Cars that drive themselves rely on GPS and online updates to know where they are and what’s happening around them. Hackers could, in theory, send fake signals to confuse the car’s location or disrupt data links.
How it’s fixed: GPS systems now include checks to spot false signals, and most data connections are encrypted so only verified systems can talk to the car.
4. Targeting the Fleet, Not the Car
When we think of hacking, we imagine one car. But many self-driving vehicles are part of fleets, managed by cloud systems that send updates and routes. If someone hacked that platform, they could, for example, flood one area with robotaxis [3].
How it’s fixed: Fleet operators use strong access controls, activity monitoring, and strict verification processes — similar to those used by banks — to keep their networks secure.
Car Hacking Examples
Year | What Happened | The Takeaway |
|---|---|---|
2015 Jeep Cherokee | Researchers remotely controlled a Jeep through its software [1]. | Proved cars need the same kind of cybersecurity as computers. |
2019 Tesla Experiment | Scientists tricked Autopilot with fake lane markings. | Taught engineers how easily sensors can be misled, and how to prevent it. |
2023 Robotaxi Gridlock | Dozens of autonomous taxis were summoned to one street [3]. | Showed that even without a hack, systems can be misused at scale. |
How Automakers Are Protecting Cars From Hackers
Carmakers and regulators have spent years preparing for these scenarios. New rules and standards mean every new model has to prove it’s built with cybersecurity in mind.
- ISO/SAE 21434: An international rulebook for how to design and test secure vehicle software [4].
- UNECE R155 and R156: Regulations that make secure software updates mandatory for new cars [5].
- NHTSA Guidelines (US): Recommendations for safe design, threat testing, and data privacy [6].
These rules don’t make cars “unhackable” — nothing is — but they do make them much, much harder to compromise.
What Drivers Can Do to Stay Safe
You don’t need to be a tech expert to protect yourself. A few small habits go a long way:
Keep your car updated. Those software updates aren’t just for convenience — they patch security issues.
Use trusted apps and accessories. Avoid third-party devices that plug into your car’s data ports.
Protect your account. If your car uses an app, set up two-factor authentication.
Pay attention to recalls. Manufacturers now send cybersecurity recalls, not just mechanical ones.
So, Should You Worry About Your Car Being Hacked?
Probably not. Car hacking makes big headlines, but it’s far from an everyday risk. Most examples you hear about are from professional researchers testing security, and their discoveries make cars safer for everyone.
The takeaway? Self-driving cars can be hacked in theory, but they’re getting smarter and safer all the time. Engineers now think about cybersecurity the same way they think about crash safety, as a built-in part of every design.
Technology will always evolve, and so will hackers, but the balance is firmly in the drivers’ favour.
References
Wired – Hackers Remotely Kill a Jeep on the Highway (2015)
ScienceDirect – Adversarial patches trick self-driving systems (2024)
The Verge – Robotaxi swarm causes traffic jam in San Francisco (2023)
ISO – ISO/SAE 21434: Road vehicles — Cybersecurity engineering (2021)
UNECE – R155 / R156 Cybersecurity and Software Update Regulations (2021)
NHTSA – Cybersecurity Best Practices for the Safety of Modern Vehicles (2022)
