By Alan Baldwin
LONDON, Feb 26 (Reuters) - Oracle and Red Bull Racing announced on Thursday a multi-year extension of their data-driven Formula One title partnership and heralded a 'ground-breaking' AI-powered strategy agent for the sport's new era starting in Australia next week.
No financial details were given for the continuation of a deal that started in 2022 and was reported at the time as being worth $300 million over five years.
"We've done some really cool stuff with AI and there's a lot more
to come," the software giant's co-CEO Clay Magouyrk told Reuters.
"We're big racing fans. We believe in F1 as a way to actually prove our technology.
"I think that having a proving ground like this is incredibly valuable as we look at how AI evolves."
ENGINE AND RULES CHANGE FOR 2026
Red Bull won constructors' titles in 2022 and 2023 and four successive drivers' championships with Max Verstappen from 2021-24.
They finished third overall in 2025, although Verstappen still won more races than anyone else, and start the season on March 8 powered for the first time by their own Red Bull Ford Powertrains engine.
The team's management also changed last year, with former boss Christian Horner and motorsport consultant Helmut Marko leaving and Frenchman Laurent Mekies taking over as principal.
Oracle Cloud Infrastructure (OCI) and Oracle AI played a big part in developing the engine -- what Mekies called a 'crazy challenge' that could have put the team at risk but now looks a solid move.
The new strategy agent, designed to work alongside human engineers and using automated data collection and interpretation of past and real-time race inputs to speed up decisions in changing conditions, could also be a significant step.
Mekies told Reuters the biggest regulation change for decades had added "quite a few dimensions to an already complex topic" with drivers having to master energy management and deploy new boost and overtake modes.
AI STRATEGIST CALLING THE RACE?
Red Bull already run billions of simulations ahead of and during race weekends and the next generation models are set to be more powerful.
Magouyrk expected to see more and more AI-driven strategy and suggested the point could come where "suddenly you could imagine that, in the same way that you have a human strategist today, you could have an AI strategist that is actually calling the race.
"Obviously the person will make the call, but you have that (AI) in real time the entire way. So it's not even just pre-simulated. It's fundamentally able to say 'this is what we should do. And here's the reasons why'.
"We're seeing that across all of knowledge work and I think it's directly applicable here. And if you do it better, the net result is that you should win more."
Mekies said the agent could also help gauge the risks associated with a driver's actions on track, particularly when an overtake might incur a penalty -- something Verstappen fell foul of notably in 2024.
"We have multiple layers of regulations: technical, sporting, financial," said the Frenchman. "And then across these regulations we have a history of live and post-race judgment...
"What we have developed with Oracle is a sort of AI agent that puts all of that together and allows us to query that agent during the race based on what we saw happening live on the track."
That could mean also deciding on whether to keep or give back a position based on "the AI agent's research on what's the likelihood of us getting away with it or getting punished for it."
(Reporting by Alan Baldwin, editing by Toby Davis)
