High-Impact IT Outages Cost Businesses $74M Annually Despite AI…

Event summary

New Relic's 2026 Observability Forecast surveyed 2,575 IT and engineering leaders across 24 countries.
High-impact IT outages cost businesses $74M annually, with a mean cost of $1.85M per hour.
36% of respondents experience high-impact outages weekly, with average detection and resolution times of 41 and 54 minutes respectively.
25% of AI agents run unmonitored in production, despite 83% of respondents agreeing AI-generated code makes observability essential.
Organizations monitoring AI agents report a 3x or greater ROI on observability investments at twice the rate of non-monitoring organizations.

The big picture

The report highlights a persistent challenge in IT operations: despite advancements in AI monitoring, high-impact outages remain frequent and costly. This tension underscores the critical need for robust observability frameworks as businesses increasingly rely on AI-driven systems. The data suggests that organizations prioritizing observability are better positioned to mitigate risks and maximize returns in the AI era.

What we're watching

AI Integration Risks: How the unmonitored deployment of AI agents will affect production stability and customer trust.
Observability ROI: Whether organizations will accelerate observability investments to capture stronger returns as AI adoption grows.
Outage Cost Trends: The pace at which high-impact outage costs will rise as AI accelerates software development and deployment cycles.