AI Ethics

When an AI system causes harm, who is liable? The maker, the deployer, or no one? Real cases are already testing this, courts are answering inconsistently, and most explanations of the problem are either academic papers or hot takes.

I built an interactive, case-driven explainer that walks through real incidents, including the Air Canada chatbot ruling where the deployer was held liable for AI misinformation, then maps the legal void: how the EU AI Act's risk tiers and the extended Product Liability Directive attempt to cover AI, where liability chains break down, and how frameworks like Microsoft's Responsible AI program assign accountability inside companies. It closes with my own verdict, including proposals like a legislated duty to signal uncertainty and mandatory named accountability chains.

A live interactive resource, built solo in React, that turns a legal grey zone into something a non-lawyer can navigate in ten minutes.

• 01

  Accountability questions become concrete the moment you follow one real case end to end.

• 02

  Building the explainer forced me to hold a position rather than just survey opinions, which is the difference between summarizing a debate and contributing to it.