Enhancing AI safety: Anthropic releases Claude 4 with increased protections

As the global race in artificial intelligence gathers pace, American AI startup Anthropic has unveiled the latest, most powerful versions of its model Claude. The company says they can write computer code by itself and play Pokemon for much longer than its predecessors.  Yuka Royer speaks with the company’s Chief Product Officer Mike Krieger about ensuring safety, fighting deepfakes and reducing AI’s environmental footprint.