AI coding assistance streamlines developers in inadvertently crafting security issues

In the rapidly evolving world of software development, the integration of Artificial Intelligence (AI) has become a common tool for developers. However, a recent study by Apiiro has shed light on a concerning trend: AI is found to multiply various types of vulnerabilities simultaneously.

Apiiro analysed code from tens of thousands of repositories and several thousand developers affiliated with Fortune 50 enterprises. The findings reveal that AI-assisted developers generate ten times more security issues compared to their unassisted counterparts.

The increased output of code by AI-assisted developers, which ranges from three to four times more than their unassisted peers, has led to a 10x increase in security issues from December 2024. This surge in security issues covers a broad set of application risks, including added open source dependencies, insecure code patterns, exposed secrets, and cloud misconfigurations.

One instance that highlights this trend is an instance where an AI-driven pull request altered an authorization header across multiple services. This alteration led to a silent authentication failure when a downstream service wasn't updated, underscoring the potential risks associated with AI-assisted development.

Code reviews, a crucial step in ensuring the security of software, have also become more complicated due to AI-assisted developers packing more code into fewer pull requests. This density makes it harder for reviewers to spot potential security issues.

The companies that have developed AI coding assistants, including market leader Checkmarx, which supports large development teams globally, are not immune to these risks. As of June 2025, AI-generated code had introduced over 10,000 new "security findings" per month in Apiiro's repository data set.

While AI offers numerous benefits to software development, it is clear that its integration also presents new challenges. As we continue to leverage AI in our development processes, it is essential to address these security concerns and find ways to mitigate the risks associated with AI-assisted development.

AI coding assistance streamlines developers in inadvertently crafting security issues