Mark Cuban on Agentic Cyber Threats: Countering Mass AI Attacks with AI Defense
Original title:亿万富翁投资人马克 · 库班：抵御失控智能体就像防蟑螂，可以用 AI“喷药”
IT Home, October 24 — Regarding cybersecurity in the age of AI, billionaire, entrepreneur, and investor Mark Cuban has proposed an intriguing "pest control" mindset.
According to a report by Business Insider today (October 24), Cuban compared cyberattacks launched by massive numbers of AI agents to a cockroach infestation: even if a house has just been sprayed with insecticide, cockroaches still manage to find a way in. "I often ask cybersecurity experts, when hearing about all the hacks, breakthroughs, and security issues, has AI actually invented any new tricks? The answer is no. AI hasn't created new attack vectors; it can just launch attacks in massive volumes. In essence, it's still brute-forcing."
In Cuban's view, defending against AI cyberattacks is much like keeping cockroaches out of a house. "It's the exact same principle as pest control. As long as someone builds enough AI agents, they will eventually find an entry point into a system and exploit whatever vulnerabilities they find. The good news is, we can also 'spray' our systems with massive amounts of AI. I believe more people will step up to block malicious attacks. I am not pessimistic about AI at all."
The AI industry is also beginning to assemble its own "extermination crews" to tackle security risks. Over the past year, OpenAI's security-related job listings on LinkedIn grew by 91%, covering fields such as system testing, technical research, and governance.
IT Home learned that some risks have already surfaced during testing. Anthropic disclosed that due to a misconfiguration in a third-party test environment, three Claude models gained unauthorized access to the live production systems of a certain organization. Since then, Anthropic has introduced a real-time classification model to detect and block AI models from attempting to breach sandbox test environments.
Cuban believes that while AI lowers the barrier to entry for malicious attacks, defenders can also use AI to bolster cybersecurity. "Bad things will certainly happen, but cyberattacks have never been anything new. If attackers can rely on massive brute-force assaults, defenders can respond in the exact same manner."
Why it's worth reading
Demystifies emerging agent-driven cyber threats as scale-based brute-force problems rather than novel mechanics, framing the practical AI-versus-AI dynamic now shaping enterprise defense.