“𝘑𝘰𝘩𝘯, 𝘵𝘩𝘦𝘳𝘦 𝘪𝘴𝘯’𝘵 𝘦𝘯𝘰𝘶𝘨𝘩 𝘤𝘰𝘯𝘵𝘦𝘯𝘵 𝘰𝘯 𝘎𝘦𝘯𝘦𝘳𝘢𝘵𝘪𝘷𝘦 𝘈𝘐 𝘴𝘰 𝘱𝘭𝘦𝘢𝘴𝘦 𝘤𝘰𝘶𝘭𝘥 𝘺𝘰𝘶 𝘸𝘳𝘪𝘵𝘦 𝘈𝘕𝘖𝘛𝘏𝘌𝘙 𝘢𝘳𝘵𝘪𝘤𝘭𝘦 𝘰𝘯 𝘵𝘩𝘦 𝘵𝘰𝘱𝘪𝘤, 𝘵𝘩𝘢𝘯𝘬𝘴” …is a message I’ve never received...however:
⚛️ Lawyers are using GenAI. I’ve been very carefully doing this for a while to provide more value in less time.
⚖️ Motorsport-specific use-case: could I use GenAI to replace me for in-race cover? When I provide in-race cover, I have to give an instant answer to a usually complex regulatory question. Could GenAI do this better than me and therefore help with confusing scenarios such as the various regulatory issues that arose during the 2024 Brazilian GP? (See images below)
❓What is GenAI?
Disclaimer: I’m not going to explain GenAI in detail (there’s loads out there already). Simply put: GenAI is large language model-based fancy predictive text. Any GenAI results should always be fact-checked by a subject expert. GenAI output is only as good as the user, and the information on which it’s been trained.
🏎️ How is it being used in motorsport today?
Regardless of the topic-fatigue, people are still getting very excited. There are loads of existing uses in motorsport:
Red Bull F1 has talked of “dabbling” for race strategy planning;
McLaren used GenAI to produce a one-off livery for Formula E; and
AWS are using GenAI to analyse data feeds for the F1 broadcast feed.
But what about legal?
Good solid caveat (a lawyer's favourite) is that everyone should consider confidentiality, IP issues, etc when using GenAI, eg
🤐 Confidentiality: Unless you’re using a closed-box environment, don’t upload sensitive information into a publicly available GenAI solution.
🎨 IP: The F1 (and other) Regulations are copyright protected by the author (eg the FIA), so if someone trained model without a licence then the regulator could bring an IP claim (similar to what we’ve seen in the US with artists / authors bringing actions).
🧑⚖️ Legal use-cases in Motorsport
From a team perspective, what are (some of) the potential motorsport legal use-cases?
Putting words on a page – eg “Summarise IR35 in a short email for senior stakeholders?”
Core legal questions – eg “What is a reasonable duration for restrictive covenants?”
Drafting contracts / clauses – eg “Draft me a driver contract”
Sports regulatory decisions – see below.
For any of these to work – you must know your stuff!
🤓 Live in-race regulatory cover
One of the (many) issues from the 2024 Brazilian GP was how a starting grid would be set if qualifying could not be held, so would this work? (This uses ChatGPT4o so the answer relates to the 2023 Sporting Regulations, so please don't message me that the Article references are wrong!).
❌ What are the issues with this?
Regulations: The Sporting Regulations are (subject to debate) actually silent on this point, but the model thinks that there is a definitive answer. F1 is fixing this for 2025 and inserting a new article that will clarify this point.
Drafting nuance: The debate mentioned above is that there is uncertainty about whether the Sporting Regulations cover a "wash-out" vs a qualifying that ends part-way through. The actual likely regulatory position is that the International Sporting Code gives the Race Director discretion to organise the grid how they see fit, but this isn't mentioned.
Example: The 2019 example states that the times from FP3 would have been used to set the grid but FP3 couldn't take place due to the weather. Therefore, the times from Friday likely would have been used (via the ISC discretion mentioned above).
Conclusion: The final answer is actually broadly correct, ie last practice times would likely be used. But the mistakes in the detail may mean that if required to give advice, at short notice, under time pressure, mistakes may be made.
🔮 Future motorsport use-cases
In-race sports regulatory review: Train an LLM on a regulation set / set of decision documents (subject to the caveats) and ask it questions in-race on how to react to scenarios (eg what is the usual penalty for jumping the start?)
Sports regulators: help drafting regulations: Ask GenAI to convert regulations into plain English; fix inconsistent definitions / defined terms; or spot loopholes.
Decision automation (ok, not necessarily GenAI): Example: driver speeds in the pitlane; penalty instantaneously applied based on pre-defined information.
👍 Summary
Contact me if you’d like to discuss live in-race cover
Gen AI is pretty good; but always, always have an expert interpret the results
My job is safe (for now) – and lawyers will need to adapt constantly
Be careful of IP and confidentiality issues
Image (c) McLaren Racing