What you need to know:
- Autonomous vehicle (AV) systems providers are using genAI to boost in-car voice assistant capabilities, reduce training costs, and improve safety and transparency.
- AV systems providers will have to build trust with regulators and prove their reliability by investing in explainable AI solutions to get fully autonomous AVs on the road.
- To stay ahead of competitors, automakers should evaluate and partner with automotive AI chip manufacturers capable of providing sufficient computing power for local LLM processing, considering factors like inference time, energy efficiency, and durability.
- Car insurance providers will need to develop new risk assessment models for fully autonomous vehicles — including provisions for cyberattack risks — and consider partnerships with automotive cybersecurity companies.
- GenAI will help turn cars into digital platforms capable of offering agentic AI features — an opportunity automakers and AV systems providers can realize by partnering with AI agent developers.
AI has always been an essential part of creating autonomous driving systems.
Now, generative AI is helping remove remaining hurdles — such as cost, explainability, and vehicle-passenger communication — that have historically slowed adoption and advancement.
GenAI is also enabling “end-to-end AI” driving systems, which use raw sensor data to directly make driving decisions, as opposed to clunkier modular approaches used in the past.
Could generative AI finally bring Level 5 (fully autonomous) vehicles onto the road? 
Source: CB Insights — Earnings transcript analysis
In this report, we look into how startups, big tech, and mobility incumbents are leveraging LLMs and other genAI advancements to:
- Reduce self-driving cars’ training costs using synthetic environments
- Improve safety and transparency for self-driving cars
- Facilitate passenger-car interactions and personalization
