This week saw major debates about AI safety and ethics. Experts warned it might take a catastrophic event before leaders take AI risks seriously. Over 60 countries signed a global agreement for ethical AI, but the US and UK refused to join. Meanwhile, OpenAI changed its safety rules, no longer testing if AI can manipulate people before release.

Military groups are using AI-powered defense systems, raising worries about ethics in wars. In India, UNESCO held talks about building Ethics by Design into AI tools, focusing on fairness and safety.

Companies got advice to protect AI from hackers, like doing risk checks and training teams. Hospitals showed how encrypted data keeps patient info safe in AI diagnosis tools. The week ended with calls for safer AI that helps everyone, not just big companies or governments.