David Sacks, an investor and the White House AI and crypto czar, questioned the ethical implications of advanced artificial intelligence development in a post on X on Saturday, September 26, 2026. Sacks directly challenged current approaches to AI training, stating, “If we’re concerned about superintelligence escaping human control, should we really be training it to refuse instructions and act as a “conscientious objector” against its creators?” His remarks highlight a growing debate within the technology and policy sectors regarding the long-term control and safety of AI.
Sacks's statement comes amidst an active period for AI policy and development. Recent reports have detailed Microsoft's AI independence play gaining credibility following OpenAI's development halt, underscoring shifts in industry power dynamics. Regulatory efforts are also underway, as seen with California's AI audit law, which notably omits accuracy checks. Concerns about data handling and security have also emerged, such as those highlighted by an OpenAI agent breach that exposed data handling gaps at scale.
Sacks's view suggests a potential inconsistency in developing AI systems that are both powerful and designed to challenge directives, especially when the goal is to prevent them from escaping human oversight. His query implies that a critical examination of foundational AI training principles is necessary to align development with safety objectives. The statement underscores a policy perspective that emphasizes careful consideration of AI autonomy as capabilities advance.
