Investing in multi-agent AI safety research

# Google DeepMind Funds Research to Make AI Systems Safer When Working Together Google DeepMind is putting $10 million toward research that tackles safety problems when multiple AI systems work together, which is becoming more common as AI gets more powerful. Right now, most safety work focuses on single AI systems, but companies are increasingly deploying many AI agents that coordinate with each other—creating new risks that researchers don't fully understand yet. This funding aims to figure out how to keep these multi-agent systems trustworthy and under control before they become widespread.