AI companies are actively flagging dangerous user searches to government authorities, even as the number of AI-related hacks continues to rise. The dual approach reflects a broader effort by tech firms to balance security risks with responsible AI deployment. The statement "Our job is to never be more than one step behind the bad guys" underscores the ongoing cat-and-mouse dynamic between AI developers and malicious actors. As generative AI tools become increasingly powerful, firms face mounting pressure to monitor for harmful use while maintaining user trust.