AI Code Risks: Security, Legal & Compliance Challenges

AI-generated code is creating new challenges for security, legal, and compliance teams. Most engineering organizations are now writing code with AI, and nearly half of them are running this AI-generated code in production. Almost every company surveyed uses AI in development, with very few planning to avoid it. What's interesting is that teams first use AI for repetitive tasks, like writing documentation or simple functions, where mistakes are small. While two-thirds of users report higher productivity and faster prototyping, error reduction is not living up to expectations. Only about a third of current users see fewer errors. The bottom line is that safeguards are piling up before AI code ships. Teams add multiple layers of protection, and no single tool makes a team immediately confident to deploy. The increased use of AI code also means more work for code reviewers, who already spend a significant part of their week on reviews. This matters because it impacts how quickly and safely new software can be developed and deployed.