Preventing artificial intelligence from triggering a world war
A mechanism for distinguishing unintended harm from deliberate attacks is urgently needed.
President Trump and Chinese President Xi Jinping discussed artificial intelligence (AI) governance during their recent Washington meeting. Both nations are intensely competing to develop more advanced AI systems and have shown no signs of slowing their progress. However, the winner of this AI race will not be immune to the significant risks AI presents.
One such risk stems from AI's growing ability to identify software vulnerabilities and execute intricate cyber operations. Since cyber warfare exhibits a pronounced imbalance between attack and defense, the nation with the most sophisticated AI models will remain susceptible to cyberattacks. This shared vulnerability encourages reciprocal restraint in targeting critical infrastructure, such as financial networks, communication systems, and power grids, using AI-enabled attacks.
Historical precedent exists for such restraint: in 2015, President Barack Obama and Xi Jinping agreed not to engage in or support cyber-enabled intellectual property theft for commercial gain, and they set up channels for information sharing, investigation assistance, and addressing malicious cyber activity. Cybersecurity firms observed a substantial drop in attacks linked to China-based groups following the agreement, although analysts pointed out that the decline had already started before the pact and might have been due to internal changes within China's military and cyber infrastructure.
Written by urgent.news from Japan Times's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
