David Sacks, an investor and the White House AI and crypto czar, stated on X (formerly Twitter) on Sunday, October 18, 2026, that current centralized approaches to AI safety in cybersecurity are counterproductive. Sacks observed that frontier models, such as OpenCode and Hugging Face, failed to analyze necessary exploits for defense because safety filters blocked payloads. He concluded, “The only durable response is decentralized: make powerful models available so defenders can build and run their own cyber defenses. The only solution to SI-powered cyberattacks is SI-powered cyber defense.”
Sacks' comments come as AI infrastructure demand sustains, with Samsung leading the memory sector upside, according to recent Gokhshtein coverage. The ongoing discussion surrounding AI's capabilities and risks has prompted figures like Mohamed El-Erian to warn of potential cracks in the AI infrastructure boom. Other market participants, including Anthony Pompliano, have suggested Bitcoin as a hedge against AI disruption, highlighting the technology's significant yet uncertain path.
Sacks' view implies that overly restrictive guardrails on AI models can inadvertently weaken cyber defenses by preventing legitimate analysis of threats, while attackers bypass such restrictions. He noted that Chinese models are only 6–9 months behind American frontier AI in cyber, operating outside U.S. jurisdiction. Therefore, Sacks believes that empowering defenders with access to powerful, decentralized AI models is crucial for effective cybersecurity in an era of sophisticated, AI-powered attacks.