New Relic Report Reveals AI-Generated Code Grades Higher in Review, Yet Triggers Rise in Production Incidents

New Relic, the Intelligent Observability Company, today released its 2026 State of AI Coding report, which reveals a central contradiction in the vibe coding era. While a staggering 94% of leaders rate AI-generated code as higher quality than human-authored code at the time of review, its deployment has triggered a significant operational tax once live. Once this code ships, 78% of respondents report more incidents, 86% report an increase in time senior staff spends fixing code and 74% report at least 25% of AI code needs significant rework when considering the past 12 months. Eighty-two percent experienced at least one production failure tied to AI-generated code in the past six months. Just 19% of organizations report no AI-generated code challenges in this time period.

Conducted in partnership with Hanover Research, the report surveyed U.S. technology leaders at upper mid-market and enterprise companies who use generative and agentic AI in software engineering on its downstream operational consequences. The data reveals a historic volume shift in software authoring that has reached well beyond startups, with 67% of technology leaders stating that AI now generates or significantly refactors between 51% and 75% of their organization's weekly code output.

"AI coding agents are no longer just autocompleting lines of text, they are driving the majority of software development across the enterprise," said New Relic Chief Technical Strategist Nic Benders. "However, our report brings to light a concerning trend: the rapid accumulation of what we're calling 'agent debt.' While leaders praise the velocity of agent-generated code during initial reviews, organizations are quietly inheriting a massive deficit of unvetted architectural logic that triggers production incidents down the line. Finding ways to mitigate agent debt is now a defining challenge for engineering organizations."

Additional key findings from the report include:

• Vibe coding clears the production bar: Vibe coding is no longer confined to sandboxes or treated as a personal productivity hack. A striking 88% of organizations have written vibe coding into formal production policies, with only 5% restricting it to non-production environments, and no respondents banning the practice outright.
• Overconfidence triggers upstream risk: The data suggests a profound misplacement of trust early in the development lifecycle. Nearly two-thirds of technology leaders (62%) report that their engineering teams often trust AI-generated code enough to ship it to production without line-by-line manual verification.
• AI-generated code scores high in initial review: A combined 94% of respondents favor AI-generated code, with 61% of leaders rating it as "somewhat higher quality" and 33% as "much higher." Only 2% of respondents perceived it as lower quality. This metric reflects subjective clarity at the time of code review rather than operational performance during live incidents.
• Observability is now table stakes: Reflecting the downstream complexity of machine-authored code, 96% of technology leaders rate observability as very or extremely important when working with AI-generated code, with not a single respondent rating it as slightly or not important.
• Telemetry moves into the AI prompt: Engineers are proactively shifting observability upstream. Nearly four in five teams (78%) now routinely prompt AI tools to include specific telemetry-such as logs, traces and metrics-directly into the generated code itself to ensure it is observable by design.

Methodology

New Relic commissioned an online survey conducted by Hanover Research in 2026, sampling 200 U.S.-based full-time technology decision-makers across IT and Engineering at upper mid-market and enterprise companies who use generative and agentic AI in software engineering. All respondents are at the manager level or above (including Directors, VPs, and C-suite executives) with meaningful software purchase authority.

Learn more

• The 2026 State of AI Coding report is now available to download.
• Read the blog post.
• Discover New Relic AI Coding Observability for production-grade monitoring of the coding phase of the software lifecycle.
• Learn how to see what your AI is actually doing at New Relic Now.