What Is “Agent Debt” and Why Is It Breaking Your AI Code?

What Is "Agent Debt" and Why Is It Breaking Your AI Code?

While AI coding tools help engineering teams ship software faster than ever, they simultaneously drive an unexpected spike in production failures.

In fact, research from New Relic reveals a striking paradox: 94% of tech leaders rate AI-generated code as higher quality during review, but 78% report seeing more incidents after deployment.

In the latest episode of the Don’t Panic, It’s Just Data podcast, host Christina Stathopoulos (Founder of Dare to Data) sat down with Nic Benders (Chief Technical Strategist at New Relic) to discuss this exact challenge: Agent Debt.

In this conversation, Nic breaks down:

The Agent Debt Paradox: Why AI-generated code looks clean and confident at review time, but creates hidden architectural flaws in production.
The SRE Dilemma: Why Site Reliability Engineers are left cleaning up AI-driven incidents—and how to fix it.
Black-Box Observability: How leading teams monitor AI workflows to stop agent debt before it impacts users.

First Name

Last Name

Company Name

Job Title

Phone

Country

Industry

Company Size

Are you currently evaluating or planning to enhance your observability, AIOps, or monitoring capabilities in the next 6–12 months?

I agree to receive marketing communications from EM360Tech and New Relic and understand I can opt out at any time.

By submitting this form, you agree that EM360Tech and New Relic may store and process your personal data to contact you with information about their products, services, and thought leadership content.

We value your privacy — your data is processed in compliance with GDPR, CCPA, and other data protection regulations. You can unsubscribe anytime using the link in our emails. For more information, please see the EM360Tech Privacy Policy.