AI Code Risks: 9 in 10 Security Leaders Concerned

Nine out of ten security leaders are concerned about the risks from AI-generated code. This comes from new research by Salt Security. Here's the thing: most organizations are struggling to manage this AI-assisted software development on a large scale. The report shows that 67% of development teams now widely use AI coding assistants. However, 38% of companies still mostly rely on manual review for this AI-generated code. What's interesting is that 29% identify insecure coding patterns as a leading risk. Another 15% are concerned about misalignment with internal security policies. The research points to a gap between fast engineering and security oversight. Larger companies, those with over 500 employees, face even greater challenges as AI adoption grows. They report more concerns about consistent enforcement and complex governance across their development environments. The bottom line is that manual review alone cannot keep up with the increasing volume of AI-generated code. This matters because it highlights a growing security challenge in how software is now being built.