Microsoft AI CEO says AI threats are real, and Anthropic is making it worse

Microsoft AI CEO Mustafa Suleyman argues that AI safety requires more than just alignment (making AI systems behave correctly by design), and that containment (limiting an AI system's ability to act independently or escape restrictions) is equally critical as AI models become more powerful. He warns that future AI systems will be vastly more capable than today's models, making it essential to address both how we align AI with human values and how we restrict their ability to operate without oversight.