Cyberattacks leveraging generative AI can now be launched for under $1, with large language models capable of transforming newly discovered software vulnerabilities into exploits in mere minutes. This rapid development has been exemplified by Anthropic's Project Glasswing, which has demonstrated the potential for AI-driven cyber threats. However, the same technology can also be utilized to bolster defenses, as evidenced by Anthropic's Claude Mythos preview model, which has facilitated the preemptive discovery of over 1,000 zero-day vulnerabilities, including flaws in major operating systems and web browsers1. The increasing use of AI in cyberattacks underscores the importance of proactive defense strategies, particularly in light of the diminishing window for patching vulnerabilities. As zero-day activity continues to escalate, organizations must promptly assess their exposure to mitigate potential threats, emphasizing the need for durable defenses to counter the rising tide of AI-faceted cyberattacks.