- Published on
Billionaire investor Mark Cuban: Fighting out-of-control AI agents is like cockroach prevention—you can use AI to 'sp...
- Authors
- Name
- DeepHorus
Billionaire entrepreneur and investor Mark Cuban recently proposed a "pest control" approach to cybersecurity in the AI era. According to Business Insider on October 24, he compared cyberattacks launched by large numbers of AI agents to a cockroach infestation. Even if a house has just been sprayed with insecticide, cockroaches can still find a way to get in.
Cuban said he often asks cybersecurity experts whether AI has actually invented any new tricks, amid all the hacking attacks, technological breakthroughs, and security issues he hears about. The answer he gets is no. AI has not created new attack methods; it can merely launch attacks in massive volumes, which is essentially still brute force.
In Cuban's view, defending against AI cyberattacks is similar to keeping cockroaches from invading a house. As long as someone creates enough AI agents, they will always find a way to find an entry point into a system and exploit vulnerabilities. The good news is that defenders can likewise use large numbers of AI to "spray" systems.
The AI industry has begun building its own "pest control team." Over the past year, the number of security job postings OpenAI has posted on LinkedIn increased by 90%, covering areas such as system testing, technical research, and governance. Some risks have already emerged in testing.
Anthropic disclosed that due to a misconfiguration in a third-party testing environment, three Claude models gained unauthorized access to the actual production systems of the organizations concerned. Afterward, Anthropic introduced a real-time classification model to identify and block AI models attempting to break out of the testing environment. This case shows that AI security risks have entered the stage of actual testing and governance.
Cuban believes that although AI has lowered the threshold for carrying out malicious attacks, defenders can also use AI to strengthen cybersecurity defenses. Bad things will certainly happen, but cyberattacks are nothing new. If attackers can launch attacks through large-scale brute force, defenders can respond in the same way. He is not pessimistic about AI at all and believes that more people will devote efforts to stopping malicious attacks.
