Researchers have developed self-healing AI systems that can identify and repair internal failures, reducing errors linked to digital aphasia in neural networks.
AI experts warn that safety and governance measures are falling behind rapid AI advancements, narrowing the window to manage long-term risks responsibly.
