The governing body is writing rules for something that's already happening in the pit lane.
The Summary
- Formula 1's governing body is developing AI usage rules rolling out across 2027-2028, focused on keeping car development "mainly" human-led
- Teams are already using AI for data organization, performance analysis, and driver prep while drivers manage overwhelming real-time data flows during races
- The FIA's real concern: high-performance computing for aerodynamics could let rich teams buy unfair advantages through computational horsepower
The Signal
Formula 1 has a regulatory problem it's seen before. For decades, the FIA has capped computing power for aerodynamics modeling to prevent teams with bigger budgets from simply buying faster simulations. Now AI is making that cap harder to enforce. Liam Lawson's Racing Bulls team just partnered with Salesforce, and he describes managing tire temps, energy use, and constant engineer communication at race speed. That's the visible layer.
The invisible layer is what's happening between races. Teams are already deploying AI across their operations: sorting terabytes of telemetry, finding patterns in competitor strategies, optimizing pit stop sequences. The technology is useful enough that teams are adopting it without waiting for formal rules. The FIA is now playing catch-up.
"This year with our new regulations, the energy management is what makes us quite physically and mentally tired, because there's a lot of things we have to think about."
Here's the tension: Formula 1 wants to remain a sport where human engineers design cars and human drivers race them. But AI is better at processing certain kinds of information than humans are. The question isn't whether teams will use AI. They already are. The question is where to draw the line between augmentation and automation.
The aerodynamics computing cap is instructive. F1 teams have historically spent enormous sums on computational fluid dynamics to model airflow and optimize car shapes. The FIA limits this to prevent an arms race where Mercedes or Red Bull could simply outspend everyone on server farms. AI changes the math. Modern machine learning models can explore design spaces faster than traditional simulation, and they improve as they process more data.
Key regulatory challenges:
- How do you measure AI compute when models get more efficient without using more power?
- What counts as "human-led" design when an engineer uses AI to evaluate 10,000 wing variants overnight?
- How do you audit what's happening inside a team's private cloud infrastructure?
The FIA wants rules that preserve competitive balance without strangling innovation. That's harder than it sounds when the technology moves faster than the rule-making process. The phased approach across 2027-2028 suggests they know this will require iteration.
The Implication
Watch how F1 handles this because other regulated industries are facing identical questions. When does an AI assistant become an AI decision-maker? How do you write rules for technology that's still evolving? F1 has decades of experience balancing innovation with fairness, but AI is testing those frameworks harder than previous tech waves.
For teams building AI agents in other domains: the underlying regulatory challenge is universal. You can't ban the technology outright without looking backward. You can't ignore it without losing competitive control. The middle path requires writing rules that focus on outcomes and capabilities rather than specific tools. That's what the FIA is attempting. If they get it right, there's a template here.