Palmer Luckey says AI cyberattacks have outrun defenses: 'time to start running'
Anduril's co-founder says outdated systems are the weak link, and the gap between AI-enabled attackers and legacy defenses is now a board-level risk.
Palmer Luckey, co-founder of defense technology company Anduril, warned that AI-powered cyberattacks are outpacing defenses and that outdated technology is leaving systems vulnerable to a new wave of AI-enabled attacks. For decision-makers, the warning reframes cybersecurity from a technical cost center into a speed problem that touches insurance pricing, disclosure obligations, and board oversight.
Palmer Luckey, the co-founder of defense technology company Anduril, has a blunt warning for anyone still treating cybersecurity as a quiet line item in the IT budget: the attackers have already moved, and the defenders have not. In comments reported by TechCrunch, Luckey said AI-powered cyberattacks are outpacing defenses, and that outdated technology is leaving systems vulnerable to a new wave of AI-enabled attacks. His conclusion was less a recommendation than a starting gun: 'It's time to start running.'
Read plainly, the warning is about a gap that already exists between how fast offensive tools are improving and how slowly most organizations replace the systems they depend on. The claim is narrow and specific: AI is accelerating attacks, defenses are behind, and the lag is being paid for in exposed systems. Everything else follows from that. The interesting question is not whether the threat is real, it is why the defense side of the ledger moves so much slower than the offense side, and what that mismatch costs the organizations caught in the middle.
Luckey is not a generic commentator on the topic. He co-founded Anduril, a defense technology company that builds autonomous systems and software for government and military customers, and he is best known publicly for founding Oculus VR, the virtual reality headset maker acquired by Facebook. That combination matters here: he sits at the intersection of two worlds that rarely share a vocabulary, the venture-backed startup scene and the slow-moving defense procurement machine. When someone from that seat says the defenders are losing ground, the audience is not just security teams. It is boards, procurement officers, and the investors funding the next generation of defense tools.
The mechanics of why AI changes the math are not mysterious. Historically, a capable cyberattack required scarce human skill: someone to write the exploit, someone to craft a convincing lure, someone to sit inside the network and adapt when the target pushed back. Generative and agentic tools compress all three. Lures can be produced at volume and in flawless local language. Code can be drafted, tested, and rewritten faster than a typical patch cycle. Reconnaissance that once took weeks of manual work can be partly automated. None of that makes defense impossible, but it changes the ratio of attacker effort to defender effort, and that ratio is what security budgets are ultimately built around.
The defender's side of the ledger is harder to move, and that is the part Luckey is pointing at. Enterprises and governments typically run a long tail of legacy systems, some of them decades old, kept alive because replacing them is expensive, disruptive, and politically thankless. Patch cycles are often measured in weeks or quarters. Public sector procurement can stretch across budget years. The result is a structural mismatch: offense iterates on the timescale of a software release, while defense iterates on the timescale of a committee. Luckey's phrase, 'time to start running,' is a way of saying that the mismatch is no longer theoretical.
The second-order effects are where this gets expensive. Cyber insurance underwriters price risk based on loss history and control quality, and a threat class that scales faster than controls do tends to push premiums and exclusions in the same direction. Regulators have moved toward faster disclosure of material incidents, which shortens the window between a breach and a public statement. Boards that once delegated the topic to a chief information security officer now find it landing on the audit committee agenda, because the question is no longer only technical, it is financial and reputational. And for the defense technology sector itself, a warning like this is also a market signal: if the gap is real, demand for tools that close it is not a cyclical bet.
For executives in adjacent roles, the practical question is not whether Luckey is right about the exact timeline, which no one can know, but whether their own organization is built for a world where the attacker's iteration speed is set by AI and the defender's is set by a change-management calendar. That is an uncomfortable question, because the honest answer for most large organizations is no. The systems that run payroll, supply chains, hospitals, and power grids were largely designed for a threat environment that no longer exists. Luckey's warning does not come with a checklist, but it does come with a deadline implied by its own logic: the gap between offense and defense is not closing on its own, and the organizations that treat it as a someday problem are the ones most likely to discover it as a today problem.
This story's Key Insights and Take-aways are locked.
Create a free account to unlock Executive Actions for one credit.
Register to UnlockAlways free for Executives Club members. Join the Club
More in Technology
Anthropic mathematician claims Claude helped crack the Riemann hypothesis
If the proof holds, AI has toppled math's most famous open problem; if it does not, it is a case study in announcing breakthroughs before anyone verifies them.
AI's $4 Trillion Wipeout Erases Nasdaq's Gains for the Year
A brutal week took roughly $4 trillion off the AI sector and left the Nasdaq flat for the year, and the damage ran the full length of the supply chain.
Anthropic mathematician says AI proved the Riemann hypothesis. Verification will take time.
A single unverified claim from inside Anthropic now sits against more than 160 years of open mathematics, and the gap between asserting a proof and surviving peer review is the entire story.


