It is 1998, during the Australian F1 Grand Prix. On lap 36, Finnish driver Mika Häkkinen pulls through the pit stop, then rejoins the race without stopping. The maneuver seems inexplicable. Häkkinen speeds back into the race, but those precious few seconds cost him dearly–losing him his lead to his teammate David Coulthard.
But this wasn’t a case of human error–something more sinister had occurred. “Someone had tapped into our radio and instructed Mika Häkkinen to enter the pits. Our radio was hacked, but we were able to manually reverse that and get Häkkinen right back on the track,” recalls Zak Brown, McLaren Racing’s CEO in a recent discussion with its cybersecurity partner Darktrace.
“Even in the days of analog, an audio breach impacted McLaren’s race. Today, a breach like that could be devastating to our race, our partners, and our economics. When it comes to security, this is why we rely on AI.”
Today’s F1 Race Cars: An Engineering Masterpiece
Behind the sleek exterior of today’s McLaren F1 cars lies a complex feat of engineering that has significantly advanced their speed, agility, and precision. However, the cutting-edge technology powering the supercars has also expanded the threat surface for hackers.
A look under the hood of a modern-day F1 car reveals a series of instruments that must all work in harmony. There are more than 25,000 separate components within the vehicle, including 11,000 components in the chassis alone, 6,000 parts in the engine, and 8,500 parts of electronics.
These components are coordinated with advanced data analytics. During a two-hour race, McLaren’s F1 Electronic Control Unit (ECU) transmits more than 750 million data points, enabling the performance of the car to be continually monitored. With more than 300 sensors, the ECU handles over 1,000 input parameters and transmits more than 300GB of live data back to the garage during an average 300 km Grand Prix.
The system is known as F1 telemetry, analyzing engine performance, the status of suspensions, gearbox data, fuel status, temperature readings, g-force measurement, and actuation of controls by the driver. Engineers at the F1 team HQ analyze this data in real time to investigate the capability and performance of their drivers and cars mid-race, including engine health, tire degradation, and fuel consumption.
This continual analysis of data enables teams to optimize performance at a granular level during the race and decide on exactly the best moment to take the car off the track. Telemetry data also helps the team decide on how they adjust the differential, the mechanism that allows the two rear wheels to rotate at different speeds–enabling significant time gains.
That’s a lot of data to protect. And many of these components are Internet-connected and therefore vulnerable to outside hackers. While legacy security tools have tried to keep attackers off computer networks, and use rule-based systems to identify known malicious behaviors, advanced threat actors–using new techniques–have triumphed.
Defending McLaren’s networks and devices from fast-moving and sophisticated malware, and other forms of attack, requires a best-in-class technological approach. Today, the indefatigable intelligence of AI is used to continually monitor the environment as a whole and works out whether digital activity is as expected, or if any element is suspect and potentially malicious.
“With the high levels of technology and connectivity we see in today’s races, McLaren finds peace of mind in the proactive and precise approach that Darktrace AI takes in securing our racing strategy, and our entire enterprise,” says Brown.
Taking the Wheel: Artificial Intelligence
The calculations that the AI makes are in real time and are constantly reassessed, as new information emerges and the context evolves–a technique called “self-learning.” This adaptiveness is critical, allowing the technology to keep pace with fast-changing, noisy data environments and protect the drivers from even the slightest disruption.
Darktrace’s cybersecurity AI was adopted by McLaren at the start of the year, and its self-learning capability–detecting brand new threat types without prior knowledge of them–proved crucial in the team’s transition to remote working during the pandemic. During this period, which saw increases in novel, COVID-19-related spear-phishing attacks, McLaren’s CIO Karen McElhatton said: “We rely on the AI to fight back against email attacks with complete autonomy and lightning speed–before damage is done…Darktrace is helping us stay abreast of a lot of the changes that are happening in the digital space, particularly at this time when we are all adjusting to the new normal.”
Lightning speed response is now essential for fighting attacks like ransomware, which can travel from an email inbox to a file share in a matter of seconds, locking down files and paralyzing operations before security teams have time to react. These fast-executing attacks, which often hit when the security team is distracted or not working, are dealt with by “Autonomous Response,” a new category of security AI that fights back against the attack, interrupting the malicious activity instantly. Powered by self-learning AI, it has a precise understanding of what is anomalous and what is not, and its interventions are surgically accurate–stopping threatening behavior, but allowing normal activity to continue.
Crossing the Finish Line, Powered by AI
Whether it’s a targeted email phishing attack aimed at a time-pressed engineer, or a vulnerability exploited in an
Internet-connected component within the car, there are many routes of attack for the adversary–yet little margin for error for the security team.
AI is supercharging McLaren’s cyber defenses by delivering autonomous threat detection and response, ensuring that emerging and novel cyber threats are dealt with rapidly, before they escalate. And it evolves with McLaren’s own organization and systems, constantly learning the “new normal.” This constant and automatic recalibration of the AI’s understanding is critical, and not just in keeping pace with the workforce–by the end of the season, an F1 race car, with its 20,000 different components, is 85 percent different from the vehicle that was originally designed and engineered.
“In F1, you pride yourself on being quick and agile, your ability to respond to a changing situation,” says McElhatton. “We needed a cybersecurity platform that could adapt with us. This is why we put Darktrace at the center of our cybersecurity strategy, which in turn, secures our racing strategy more broadly.”
All technologies employed by the McLaren F1 team serve one purpose–to win the race. But as F1 cars become more advanced, sophisticated hackers will look to take advantage and derail races instead. With AI, cybersecurity has upped its game, fighting the hacker with precision from inside the system–and keeping organizations like McLaren on track.
For more great content like this check out:
The Next Normal: Remote Working Revealed New Security Holes—Returning To The Office Won’t Close Them
Finger on the Pulse: How AI Is Helping Businesses Adapt to the New Normal
The Code War: Cyberattacks Are Redrawing the Battle Lines of Global Conflict
A New Class of Threat: Educational Institutions Find Themselves On the Front Lines of Cyberattacks
AI Will Be a Crucial Tool in the Fight Against Next-generation Security Threats
Only as Strong as the Weakest Link: How Cyber AI Protects Global Supply Chains
The Reeducation of AI: A Self-Learning Approach
Top Experts: Pandemic Has “Exponentially Expanded” Corporate Security Vulnerabilities
AI in Healthcare: Protecting the Systems that Protect Us
AI: Enforcing Normal In Extraordinary Times
How AI Is Future-Proofing the Cities of Tomorrow
Offensive AI: Surfacing Truth in the Age of Digital Fakes
How AI Battles Security Threats without Humans
Mimicking a Cybersecurity Analyst’s Intuition with AI
This story was produced by WIRED Brand Lab for Darktrace.
