Artificial intelligence is fascinating, transformative and increasingly woven into how we learn, work and make decisions.

But for every example of innovation and efficiency — such as the custom AI assistant recently developed by an accounting professor at the Université du Québec à Montréal — there’s another that underscores the need for oversight, literacy and regulation that can keep pace with the technology and protect the public.

A recent case in Montréal illustrates this tension. A Québec man was fined $5,000 after submitting “cited expert quotes and jurisprudence that don’t exist” to defend himself in court. It was the first ruling of its kind in the province, though similar cases have occurred in other countries.

AI can democratize access to learning, knowledge and even justice. Yet without ethical guardrails, proper training, expertise and basic literacy, the very tools designed to empower people can just as easily undermine trust and backfire.

Guardrails are the systems, norms and checks that ensure artificial intelligence is used safely, fairly and transparently. They allow innovation to flourish while preventing chaos and harm.

The European Union became the first major jurisdiction to adopt a comprehensive framework for regulating AI with the EU Artificial Intelligence Act, which came into force in August 2024. The law divides AI systems into risk-based categories and rolls out rules in phases to give organizations time to prepare for compliance.

The act makes some uses of AI unacceptable. These........

© The Conversation