CodeRabbit, a company that builds AI-assisted code review tools, is using OpenAI's GPT-6.1 Sol model to investigate security risks across large and complex codebases. The work centres on a security agent that scans code for potential vulnerabilities and presents its findings to developers. The agent is designed to help engineers work through what it finds. Developers can review each finding and the fixes the system proposes, rather than receiving a raw list of alerts to triage on their own. The show notes describe this as a way to strengthen security in codebases that are too large or tangled for manual review to cover reliably. For Australian businesses and professionals, this changes little on its own. Teams that already use AI code review may find it a useful benchmark for security tooling, but any adoption carries licensing costs and requires developers to check the proposed fixes before merging them. If the code being scanned contains personal information, the Privacy Act 1988 applies to how that data is handled, so teams should confirm where the code is processed and what the vendor retains. No new regulation is involved.