- “AI fixes the writing errors but creates time bombs,” says APIIRO
- 10x More security results were highlighted in June 2025 against December 2024
- AI is 4x faster than humans, and can also reduce syntax failure and logical errors
New research has claimed that despite its promised progress in helping code faster and more effectively, the use of AI tools for coding may possibly throw security issues everywhere.
APIIRO has quantified the vulnerabilities that companies could expose to using AI-generated code, find AI-assisted developers were able to write 3-4x more code than their friends, but their code introduced 10x more security problems.
The vulnerabilities varied in severity, and although they were not always utilizable errors, the code quality still saw a noticeable fall.
AI-Generated Code can be really dirty
APIIRO found uncertain patterns, exposed secrets, new addictions and Sky-forced configuration across AI-generated code as well as the utilized errors that pose the greatest risk.
In June 2025, AI-generated code threw 10,000 new security results per year. Month, a ten times increase in the six months from December 2024.
However, artificial intelligence also has many clear advantages. According to the data, syntax error fell in AI-written code by 76%and logical errors fell by over 60%. Conversely, privileges increased – 322% and architectural design errors increased by 153% – questions that reviewers are struggling to spot.
“AI fixes the writing errors, but creates Timebombs,” the researchers summarized.
APIIRO also noticed AI-assisted developers exposed sensitive keys almost twice as often as their unassisted comrades. “Because assistants generate major changes with multiple files, a single identification can be propagated across multiple services or configurations before anyone notices,” explains explains.
All this in an era where AI coding actually gets a mandate – not only supported – by companies such as Coinbase and Citi. Large tech leaders also indicate that about a third or more of their new code is AI-generated.
Companies should consider implementing additional protective measures when using AI-generated code as it also acts as a reminder that human supervision, logic and experience cannot be overlooked.
