According to Google DeepMind, they have updated their Frontier Safety Framework, a set of protocols designed to mitigate severe risks associated with AI development, as part of their commitment to balancing innovation with safety in the progress towards Artificial General Intelligence (AGI).