The rise of artificial intelligence technology has been a double-edged sword in the fight against criminal cyberattacks. The rapid development and proliferation of faster, more efficient, and agentic programs have slowed ransomware and phishing attacks. Simultaneously, white-hat AI tools that hunt for zero-day vulnerabilities have reduced human error in overall system design. Cybercrime, however, has continued to grow more potent. Criminals now use AI to research their targets and craft convincing phishing attempts, and bad actors are finding zero-day vulnerabilities of their own.  As new technology emerges, these fights are less direct confrontation than a race against the clock. When defenders lose that race, the results are disastrous.

Our adversaries have put immense pressure on American infrastructure, endangering our economy and welfare. According to the FBI, Americans lose nearly $21 billion in tangible finances annually to cybercrime. Losses on this scale are unsustainable. The question is not whether this is a threat, but how to respond. As the federal government builds its capacity to go after cybercriminals, something must close the gap in the meantime. Bureaucracy is slow by design. That serves good governance, but it is a liability against criminals who face no such constraints.