At race speed, Liam Lawson is listening to his engineer, watching tire temperatures, managing energy use, and making decisions that can affect lap time. In Formula 1, the flow of data has become part of the driver’s workload. AI is now changing how quickly teams can process that data and turn it into usable information.
“This year with our new regulations, the energy management that we have to look after is what makes us quite physically [and] mentally tired, because there’s a lot of things we have to think about,” says Lawson, whose Visa Cash App Racing Bulls F1 team recently announced a new technology partnership with Salesforce. “There’s a lot of things that I have to manage while I’m driving and make sure that I do consistently.”
That includes tire temperatures, how much energy is used during a lap, and constant communication between the driver and pit wall.
“[There’s] a lot of communication with the team as well,” Lawson continues. “My engineer talks to me every single lap, and that’s a lot of stuff that we’re communicating and processing as well. For us, it’s all in the car.”
Formula 1 teams have always looked for even the slightest competitive advantages through engineering, strategy, and data. AI gives them another way to find those advantages. Teams are using AI tools to organize information, analyze performance, support engineering work, and prepare drivers. The technology is already affecting how teams work between races and during race weekends.
The FIA, Formula 1’s international governing body, is now developing rules for how AI can be used in the sport. Updates are expected to phase in across 2027 and 2028, with an emphasis on keeping car development led mainly by human engineers.
Much of the concern involves high-performance computing. The FIA has long regulated the computing power teams use for aerodynamics modeling, which helps determine car shape. Those rules were designed to limit a spending race in which teams could gain an advantage by investing heavily in faster and more powerful simulation tools. AI and machine learning raise similar concerns.
“The same is true with AI and ML,” says Dominic Harlow, FIA Single Seater Deputy Technical Director. “ We think that there’s a legitimate risk if we just left it completely alone, it would become a huge area of development, potential spending, and maybe [a] performance differentiation that detracts ultimately from the racing.”
The FIA’s goal is to allow useful technical development while preserving the human work that remains central to Formula 1.
“We also want to preserve some of the intellectual challenge for the teams, for the engineers, as well as for the drivers,” Harlow says. “F1’s a team sport. There’s driver skill, but there’s engineer[ing] skill, there’s operational skill, there’s even the people putting together the wind tunnel models in the aerodynamics department, the people designing new shapes to evaluate. They’re all putting their creativity and blood, sweat, and tears into doing it.”
Teams are already working with AI inside those boundaries. At Atlassian Williams, Russell Paddon serves as Driver Performance Engineer, a role that connects the team’s drivers with its engineering operation. He helps translate driver feedback into engineering requirements that can improve performance on track.
Paddon says he uses AI tools across Atlassian and the team’s latest AI partner, Anthropic, to make that work faster.
“The benefits we’ve had from some of the Atlassian toolsets [Rovo] has been more on organization and structure and how we share information internally,” he says.
In Formula 1, saving time can matter as much as adding technical capacity. Teams have limited windows between races and between sessions during a race weekend. AI can help reduce the time spent sorting notes, searching documents, and preparing information for engineers and drivers.
“ Often, time is our biggest enemy in Formula 1 because the season moves on so quickly,” Paddon says. “We’re always struggling to stay ahead of the curve and in between races trying to learn [from] the last weekend and then take it forward to the following weekend is really quite important. That’s where AI has kind of shortened that timeframe.”
For Paddon, who joined Williams at the end of the 2024 season, one of the most useful applications has been historical analysis. Instead of manually searching through old reports and studies, he can query the team’s existing knowledge base for relevant patterns.
“Within the Atlassian suite of tools, [I can] query [a] database,” Paddon says. “For example, I might want to look at a trend in driver performance, which was two years ago. It can dig through all the information we have and pull out any relevant projects, reports, studies, any analysis that’s been done, and [it] just brings it all into one convenient place. So, that’s been a big game changer.”
AI has also changed how he handles technical questions. In the past, answering a new performance question often meant writing code or building a tool for a specific use case. Paddon says that work now takes less time.
“A lot of my time was previously spent either having to code and write tools to be able to answer some questions, and I’d spend probably 60% of the time actually trying to do that and get the data that I was interested in and not really adding value to the team because I’m just having to keep writing new scripts every weekend for a different use case,” Paddon admits. “[I was] not really adding value to the team because I’m just having to keep writing new scripts every weekend for a different use case. That’s completely [almost] disappeared in my experience, where I can very quickly state what I need from a dataset, and I can query it. I can [then] get Claude to either write me a tool or I can very quickly just ask it to return the results.”
That gives Paddon more time to focus on what the data means for the driver, the engineers, and the car. He says the tools have improved his relationship with drivers and helped him collaborate more closely with other engineers, including on car development.
Drivers are also using AI outside the garage. Lewis Hamilton recently appeared in a campaign for Perplexity Computer that showed him using the tool before a race to create an interactive box-breathing guide and curate a music playlist.
“Perplexity is a product that helps him find those extra percentage point improvements to help him be the best in every part of his life,” says Ryan Foutty, Perplexity’s VP of Business. “He’s also separately talked about publicly how he’s using it around his training and running: what should he eat before and after [and] his recovery plan. What you see in the content, it ultimately comes from real use cases that Lewis has had.”
Lawson, who is also a music fan, says he uses AI tools to refresh his playlists and learn how to play music. “It’s just something that’s a lot easier now than it was probably 10 years ago,” he says. “For me, music’s a way to switch off, reset, and unwind more than anything.”
These uses vary widely. Some help teams organize internal information. Some support technical analysis. Some help drivers prepare or unwind. Together, they show why the FIA is trying to define AI’s role before it becomes harder to govern.
As teams adopt more AI tools, Formula 1 will need to decide which uses fit the sport’s competitive structure and which could distort it. Teams will keep looking for small gains. The FIA is trying to preserve the role of human skill, engineering judgment, and operational decision-making.
“It’s just the philosophy of maintaining a sporting element rather than machines racing machines,” says FIA’s Harlow.