November 20, 2026

Speed without guardrails: LaunchDarkly report exposes risks of uncontrolled use of AI in software delivery 

Person using laptop with ai integration logo displayed.

Today, LaunchDarkly, the global software feature management platform, releases a new report, The AI Control Gap, revealing that as AI has made software delivery faster than ever, it has also created new risk and pressure for developer teams.  

 

AI is producing faster code, but with less confidence 

  

The report highlights that nearly all teams (94%) report that AI has accelerated code development. However, 91% say their teams have become more cautious about pushing those AI-accelerated changes into production. 

  

Faster delivery is creating more pressure and risk for teams 

 

81% of respondents say their teams knowingly shipped risky code in the past six months due to deadline pressures. This has created a cycle where 38% of teams spend more than a quarter of their time resolving incidents.  

 

The most significant impacts to software delivery include: 

  • 42% of respondents say their teams are reporting lower morale and higher burnout.  
  • 99% say they have runtime guardrails in place, yet approximately 70% still roll back at least weekly. 
  • 69% of teams hotfix or roll back weekly or more.  
  • 83% say that releases containing AI-generated code are somewhat or very likely to cause production issues. 

  

Notably, more than 60% of organizations report customer-visible issues such as performance degradation, missed SLAs, or churn. 45% of respondents also reported that slow or risky releases led to lost market opportunities. 

   

Joe Byrne, Global Field CTO at LaunchDarkly, said, “Our research reveals the AI Control Gap: the gap between how fast AI enables us to ship code and how safely we can deploy it in production. The tension between speed and control is becoming one of the biggest challenges in modern software delivery. 

 

“Success shouldn’t just be measured by how fast teams ship, but how safely they do it. As 77% of teams still struggle with the pace and risk of AI-accelerated development, the business case for resilience is growing, and enterprises must lead by making controlled release, progressive delivery, and runtime stability core business metrics, alongside speed and cost.”