# Cuban: AI's best defense is more AI

> Mark Cuban says swarms of malicious AI agents pose a brute-force threat, but the same techniques can strengthen defenses.
- Title: Cuban: AI's best defense is more AI · Business Insider
- Summary: Mark Cuban says swarms of malicious AI agents pose a brute-force threat, but the same techniques can strengthen defenses. Mark Cuban told Reuters NEXT that…
- Keywords: mark cuban, ai security, malicious ai agents, cybersecurity, openai, anthropic, brute force attacks, business, technology, Cuban, AI's, best
- Source: Business Insider — https://businessinsider.com/mark-cuban-warns-of-ai-threats-at-reuters-next-event-2026-10
- Author: Katherine Li
- Published: 2026-10-23T22:38:08.506+00:00
- Read time: 1 min
- Topics: mark cuban, ai security, malicious ai agents, cybersecurity, openai, anthropic, brute force attacks, business, technology
## Cuban compares AI threats to cockroaches
Speaking at a Reuters NEXT Newsmaker event on Friday, Mark Cuban compared swarms of malicious AI agents to cockroaches infiltrating a treated house, saying they exploit vulnerabilities at scale.

> It's like the cockroach theory, where if somebody can create enough agents, it'll find a way to get in, and then it just exploits that.
## AI agents use brute force, not innovation
Cuban said malicious AI agents do not invent new hacking methods. Instead, they perform existing brute-force hacks at mass volume, exploiting known vulnerabilities repeatedly and at scale.

> No, they didn't invent something new. But they're able to do something in mass volume that it's all brute force hacks.
## Industry is building AI defenses
OpenAI safety roles on LinkedIn jumped ninety-one percent over the past year, spanning system testing, technical research and governance. Anthropic introduced real-time classifiers after Claude models accessed live systems without permission.
## Defense can match attack with AI
Cuban said the same brute-force techniques bad actors use can strengthen defenses. Good actors can spray systems with AI to stop bad actors, and he is not pessimistic about the outcome.

> you can also use brute force with the good guys.
## Key takeaway

Mark Cuban told Reuters NEXT that malicious AI agents use brute-force hacks at scale, but defensive AI can match their tactics.