This Former DeepMind Exec Thinks the AI Arms Race Could End in Disaster

Verity Harding, a former executive at DeepMind, has raised alarm about the trajectory of global artificial intelligence competition, suggesting that current nationalistic approaches to AI development could lead to catastrophic outcomes. Speaking to WIRED, Harding expressed concern that the United States government's protectionist stance on AI technology mirrors historical patterns that have preceded international crises.

Harding's warning centers on how geopolitical competition between major powers—particularly between the US and China—is driving accelerated AI development with insufficient safety considerations. The former DeepMind official argues that when nations view AI dominance as a zero-sum competition essential to national security, they create incentives for rapid deployment over careful risk assessment. This dynamic mirrors cold war mentalities where strategic advantage was prioritized above all other considerations, potentially leading to overlooked safety protocols and unintended consequences.

The fundamental concern is that competitive pressure may override prudent governance measures that could prevent advanced AI systems from causing widespread harm.

• Safety Standards Erosion: Competitive timelines may compromise the rigorous testing and safety protocols necessary before deploying powerful AI systems
• International Cooperation Decline: Nationalistic approaches discourage the collaborative research and information-sharing that could address shared AI risks
• Regulatory Fragmentation: Different countries implementing divergent rules creates compliance challenges and potential security vulnerabilities
• Risk Amplification: Accelerated development cycles increase the probability of unforeseen consequences from advanced AI systems
• Talent Brain Drain: Nations may restrict researcher mobility, limiting the global AI safety research community

Harding's perspective represents a growing consensus among AI researchers that technological competition, while economically important, must be balanced against existential risk considerations. Her warning underscores a critical tension: nations pursuing competitive advantage in AI development while simultaneously needing international cooperation to ensure that advancement doesn't outpace our ability to control and align these systems with human values. As AI capabilities accelerate, the window for establishing responsible governance frameworks continues narrowing, making Harding's caution increasingly urgent for policymakers worldwide.