When your AI system makes decisions that affect human safety, who bears the moral responsibility?
The problem
Autonomous and semi-autonomous systems in the automotive sector operate in conditions where moral decisions happen in milliseconds. Who bears responsibility when the system causes harm? How do you document the ethical choices embedded in system design? How do you demonstrate, to regulators and to the public, that those choices were made deliberately rather than by default?
Most organisations in this sector have invested heavily in technical safety. Fewer have invested equally in the ethical architecture that gives those safety decisions meaning and accountability.
Our approach
We apply Moral Situation Modelling to map the ethically relevant dimensions of the system’s lifecycle at different levels of granularity: the points where human oversight is required, where accountability must be allocated, and where the system’s behaviour reflects choices that need to be made explicit. This work precedes and informs the Ethics-Based Audit, which examines whether those choices were implemented as intended.
What structured ethics governance produces
Organisations that complete this process have a documented ethical architecture for their AI systems. They can demonstrate to regulators, partners, and insurers that the moral dimensions of their systems were addressed by design, not discovered after an incident.