Evidence suggests that while OpenAI is directing compute power toward discovering vulnerabilities in open-source software like Linux and Firefox, the scale of this investment is relatively small compared to other AI research goals. Cybersecurity professionals argue that the current pace of defense is struggling to keep up with the mountain of work generated by the accessibility of powerful AI-assisted exploitation tools.