Eric Schmidt Warns Against AI Arms Race, Suggests Risky Alternative

Image Credentials: Generated by Open Chronicle AI DALL·E

By Staff Writer with Agencies

Washington, D.C. – Former Google CEO Eric Schmidt has issued a stark warning against a U.S.-led push for artificial general intelligence (AGI), arguing that a “Manhattan Project” approach could escalate tensions with China and destabilize global security. His recently co-authored paper, Superintelligence Strategy, suggests that an aggressive pursuit of AI supremacy could provoke hostile countermeasures, similar to Cold War nuclear standoffs.

Schmidt, alongside co-authors Alexandr Wang of Scale AI and Dan Hendrycks of the Center for AI Safety, warns that if the U.S. attempts to dominate AGI development, rivals may not passively accept the imbalance. Instead, China and other nations could retaliate, triggering a dangerous AI arms race.

A New Strategy: Deterrence Over Domination

Rather than a unilateral race for AGI, Schmidt and his co-authors propose a deterrence-based approach called Mutual Assured AI Malfunction (MAIM). This strategy involves:

• Expanding cyber capabilities to disable adversaries’ dangerous AI projects.
• Restricting access to advanced AI chips and open-source AI models.
• Strengthening domestic AI production to prevent reliance on foreign tech.

This marks a shift in Schmidt’s AI stance. Previously, he championed aggressive competition with China. Now, he argues that a more defensive approach could prevent AGI from becoming a destabilizing global force.

The Risks of AI Sabotage

While the MAIM strategy aims to prevent an uncontrolled AI arms race, critics argue that sabotaging other nations’ AI efforts could backfire. China, with its advanced cybersecurity and AI capabilities, may respond by hardening its systems against U.S. interference or escalating retaliatory measures.

Additionally, with multiple global powers pursuing AGI, a deterrence-first strategy may not be enough. If AI development continues across various countries, the risk remains that AGI could be weaponized by any one of them, despite U.S. efforts to prevent it.

As world leaders navigate the complexities of AI policy, Schmidt’s proposal raises crucial questions: Should the U.S. aim for dominance in AGI, or should it prioritize preventing an uncontrolled global arms race? With the stakes higher than ever, the future of AI governance remains uncertain.