AI Security
Luckey: AI-driven hacking is outrunning defensive systems—organizations must speed up security upgrades
At the Times DealBook Summit, the Anduril Industries co-founder warned that adversaries are exploiting outdated defenses faster than organizations can patch them.
TechCrunch — AI
Palmer Luckey, co-founder of defense-focused Anduril Industries and former Oculus creator, raised alarms this week about a growing security imbalance: artificial intelligence is enabling attackers to move faster than many organizations can defend themselves.
Speaking at the New York Times DealBook Summit on Tuesday, Luckey pointed to the recent wave of system failures as a symptom of a deeper problem—not simply careless patching, but legacy infrastructure struggling against next-generation threats.
The AI Attack Advantage#
Luckey emphasized that adversaries are actively weaponizing AI-generated exploits, allowing them to discover and deploy vulnerabilities at an accelerating pace. The mismatch between cutting-edge attack methods and outdated defensive systems has created an opening that hostile actors are actively exploiting.
He singled out North Korea as particularly aggressive, noting that attacks are being pursued for both geopolitical leverage and financial gain. The situation was underscored when U.S. Vice President J.D. Vance canceled his conference appearance to attend security meetings instead.
The Defense Lag#
Luckey acknowledged the uncomfortable reality: "Everything can be hacked" for now, particularly with emerging techniques on the horizon. The critical issue is not that defense is impossible, but that established systems cannot evolve quickly enough to match the speed of AI-driven attack innovation.
Organizations clinging to outdated platforms are, in effect, frozen targets. As Luckey framed it, preparation and investment in modern security infrastructure will ultimately determine which entities weather this wave.
A Path Forward#
Despite the grim near-term outlook, Luckey argued that AI itself may eventually tip the scales in defenders' favor—but only if they embrace it as aggressively as attackers have. The institutions and nations that move fastest to adopt AI-powered defense will gain the decisive edge.
His closing message was blunt: "It's time to start running."

