Lawyers can help make AI safe.

Legal guidance on AI typically focuses on AI as a productivity tool, but AI risk is more than hallucinations. Risk arises in the training of AI models, the architecture of AI systems, their integration with your own workflows, and their ongoing use.

AI risk isn’t static. A model’s weights may remain fixed, but what a system can do (and the risks it poses) can change as it gains access to new tools, data and environments. Lawyers need to understand these dynamics across the AI lifecycle: advising on procurement, scrutinising safety and performance claims, designing effective oversight, and advising when systems cause harm.

Lawyers have a vital role in making AI safer. Fulfilling it requires a deeper understanding of the technology.

The legal profession’s response to AI must extend beyond its own use of the technology to how AI is built, governed, and held to account across society.

Lawyers already help shape the rules, incentives, evidence, and decisions that determine how the risk is handled. Yet AI risk remains absent from most legal training and continuing professional development.