OSFI E-23 sets model risk expectations for Canada’s banks and insurers, including credit algorithms, actuarial tools, fraud systems, and language-model applications.
In January 2026, Character.AI and Google settled five wrongful death lawsuits. OpenAI is facing eight more, with estimated damages per case ranging from $20M–$100M. Every single one of these tragedies was preventable. We mapped each lawsuit to its root cause, and the testing capability that would have prevented it. Here’s what we found:
