Overview
- Nvidia’s Chief Security Officer David Reber Jr. declared in an Aug. 5 blog post that its AI GPUs “do not and should not have kill switches and backdoors.”
- China’s Cyberspace Administration summoned the company in late July to demand documentation on alleged vulnerabilities in its H20 AI chip.
- Nvidia warned that embedding remote-shutdown or tracking mechanisms would undermine global digital infrastructure and fracture trust in U.S. technology.
- The blog post cited the Clipper Chip debacle of the 1990s to illustrate the risks of government-mandated hardware backdoors.
- U.S. lawmakers are debating the Chip Security Act, which would require exported AI chips to include location-verification technology.