AI Code Security: Developers Skip Vital Steps

Nearly all developers are using artificial intelligence to write code, but very few are securing that code as they go. Only 18% of developers apply security continuously while writing code, despite 96% using AI tools. Here's the thing: companies where 81% to 100% of their production is AI-generated code are nearly three times more likely to ship known vulnerabilities. This is compared to those using AI for just 1% to 20% of their code. What's interesting is that 75% of organizations knowingly deploy vulnerable code at some point. Also, 95% of security leaders feel pressure to delay compliance issues due to business deadlines. Despite this, 93% of organizations acknowledge a recent breach tied to their own applications, even though 73% describe their security as advanced. The bottom line: 78% of organizations still lack formal AI governance policies. This highlights a significant risk for anyone using software developed with AI.