Formula 1’s rule-makers face a demanding engineering problem: anticipating how teams might exploit technical regulations before the cars reach the track. For IT leaders, the challenge is familiar: expanding what a specialist team can evaluate without adding people at the same rate.
On Oct. 28, the Fédération Internationale de l’Automobile (FIA) and Amazon Web Services announced a global partnership covering AI and cloud infrastructure. The agreement combines engineering workloads with digital licensing and business application modernization, giving IT leaders several practical use cases to examine.
The deal covers three areas: high-performance computing and AI for aerodynamic simulations supporting future single-seater regulations, an expansion of the FIA’s Digital Licensing Platform, and a broader migration of FIA business applications and hosted services to AWS over the next several years.
Here are my five takeaways that should matter to IT leaders.
Elastic compute is the great equalizer
AWS’s separate partnership with Formula 1 began in 2018. In a 2023 report on F1’s work developing aerodynamic regulations, Autosport reported that a single high-fidelity computational fluid dynamics run had fallen from about 60 hours to roughly 10 hours on AWS, with multiple jobs able to run in parallel.
The practical lesson is that on-demand compute can help a specialist team evaluate more scenarios without owning infrastructure sized for peak demand. Most enterprise IT teams will never have the headcount of their largest competitors. On-demand compute can reduce capacity constraints, although its value depends on workload performance and total cost. It turns a capacity problem into a scheduling and spending decision.
AI should sit on top of your simulation, not replace it
Dominic Harlow, the FIA’s deputy technical director for F1, told The Race that its core workflow uses OpenFOAM for computational fluid dynamics. He described exploring surrogate models trained on CFD results to accelerate parts of that workflow.
Such models approximate simulation outcomes rather than solving the underlying physics for every new case. The approach remains exploratory; the reporting does not establish a deployed FIA model or a validated performance benchmark.
For IT leaders, the useful lesson is to assess whether AI predictions remain accurate across the conditions where they will be used. Training data quality matters, but engineers also need comparison tests, acceptable error thresholds, and a clear route back to the established simulation process.
Keep humans in the loop, especially when the stakes are regulatory
The FIA is being deliberately cautious. Generative AI will not yet be used to propose design ideas, and Harlow described the work as still exploratory and human-driven. The Race also identifies potential uses for large language models in supporting teams’ regulatory inquiries and stress-testing draft regulations, without specifying a deployment timetable.
That’s smart sequencing, not a lack of ambition. When an output becomes a binding rule that affects billions of dollars in team investment and driver safety, explainability matters more than speed. IT leaders deploying AI in regulated or high-consequence workflows should establish clear ownership and accountability, start with AI that augments expert review, validate its accuracy against known outcomes, and only then expand its authority. “Red team the rulebook” is also a strong use case for any organization drafting policies, contracts, or compliance controls.
The unglamorous back office is where transformation pays off
The aerodynamics and race analytics get the headlines, but the Digital Licensing Platform is the bigger story. The FIA has 246 member organizations across five continents. The platform, unveiled in 2025 and first rolled out in Canada and Sri Lanka, handles applications through approval on web and mobile. It’s now expanding to more countries with multilingual support, FIA University integration, and license renewals, with event management for volunteers and officials coming next.
The FIA says national sporting authorities retain full sovereignty over their data and decision-making. However, its public announcement does not detail the platform’s architecture or data residency arrangements. Multinational IT teams considering a similar platform should verify hosting locations, access permissions, and data export rights.
FIA President Mohammed Ben Sulayem said the partnership would make participation easier, help member clubs work more efficiently, and support the ambition to double global motorsport participation.
The metric he chose is notable because it’s not cost savings or uptime, but participation. Tying digital transformation to a measurable business outcome, rather than an IT metric, helps projects keep funding beyond year one.
Strategic partnerships bring leverage, and dependency
Kristin Shaff, global director of events, sports and entertainment at AWS, described motorsport as a demanding environment rich in data. AWS gets a high-profile showcase, while the FIA gains expanded access to AWS technology.
IT leaders should weigh both sides of that trade-off. The FIA has said that more business applications and hosted services will move to AWS over the coming years, deepening that dependency. There’s also an obvious governance question. AWS is a partner of Formula 1 and now the rule-maker’s AI and cloud provider. The public announcement does not detail how confidential workloads are isolated or who can access them. IT teams introducing AI into these environments should also define boundaries around the credentials, tools, and data it can access. OpenFOAM’s open-source foundation offers flexibility, but the portability of the complete simulation and AI workflow has not been established.
What this means for IT pros
The lessons here are enterprise lessons. IT leaders should:
- Use elastic compute to punch above your weight. Identify the workloads where a capacity constraint, not a talent constraint, is holding the business back, and fix those first.
- Treat your data as the product. Surrogate models are only as good as the simulations and history they’re trained on. Invest in data quality before model selection.
- Sequence AI by consequence. Start with augmentation and review in high-stakes workflows, prove accuracy, then expand autonomy.
- Define local control on shared platforms. Specify which data and decisions business units or regions control, and verify the permissions and contractual terms supporting that control.
- Tie the transformation to a business outcome. Since the FIA chose participation, pick a metric your board cares about.
- Keep the core portable. Open-source and open-standard foundations give you leverage when it’s time to renegotiate a strategic partnership.
The FIA will always be outnumbered by the teams it regulates. The AWS partnership changes how much of that gap compute and AI can close. Most IT organizations face the same challenge against larger competitors, and the FIA’s approach is a useful model.
Read more: Before expanding AI’s role in business workflows, examine why approving a tool still requires clear controls over data access and accountability.