TEL AVIV, Israel and SAN FRANCISCO, Nov. 04, 2026 (GLOBE NEWSWIRE) -- Komodor, the Agentic Operations Platform company, today released its 2026 Agentic Operations Survey, revealing that engineering organizations are rapidly expanding their use of AI agents but remain cautious about granting them greater autonomy without adequate governance, reliable performance and cost controls.
The survey found that building safe guardrails is the most frequently cited challenge (53.6%), followed by governance and security (50.4%) and standardization and accuracy (45.9%). Meanwhile, nearly 84% of respondents report running fleets of 10 or more operational agents, underscoring the growing need for a shared foundation to manage agentic fleets safely and efficiently at scale.
The survey of 1,000 engineering and technology leaders also found that 75.7% maintain some restrictions on autonomous operations, constraining agents to read-only tasks, requiring human approval for changes, or limiting automated actions to noncritical workloads. Meanwhile, production incident investigation (43.2%) and remediation (44.2%) are nearly tied among leading AI agent use cases, demonstrating growing demand for automation that can both diagnose and resolve production problems.
The findings highlight a growing concern: organizations are expanding AI automation across critical infrastructure faster than they are establishing consistent policies and technologies to control what agents can access, change, and execute.
Read the full 2026 Agentic Operations Survey here: https://komodor.com/resources/komodor-2026-agentic-operations-survey-report/
“We’ve seen this pattern before when enterprises began scaling microservices. Building one agent is straightforward, but operating hundreds across production environments introduces an entirely different set of challenges,” said Ben Ofiri, co-founder and CEO of Komodor. “But there’s a critical difference this time: agents can make decisions and take actions on their own. What’s needed is a shared foundation that gives agents the operational context to make informed decisions, enforces appropriate controls, and continuously evaluates their performance as they take on more responsibility.”
Key Findings from the Survey
Beyond the growing autonomy of production AI agents, the research identified four significant trends:
- Governance and control remain top priorities: Building safe guardrails is the most frequently cited challenge (53.6%), followed by governance and security (50.4%). Another 42.7% identify insufficient visibility into agent permissions. The findings underscore the need for engineering organizations to control what agents can access and execute while ensuring their decisions remain reliable and consistent across production workflows.
- Ensuring reliable AI agent outcomes remains a major challenge: Nearly half of respondents (45.9%) rank standardization among their top three challenges, illustrating the need to ensure agents produce consistent, accurate results. Continuous evaluation against real operational scenarios, testing changes to models and prompts, and using performance feedback to improve agents over time are essential as organizations expand autonomous operations.
- AI-driven development puts greater pressure on production operations: As coding agents accelerate software development, SRE, DevOps and platform engineering teams face growing demands to manage deployments, investigate incidents and resolve production problems. The survey reflects this shift, with routine task automation (49.1%), support ticket triage (47.1%), self-remediation and AI SRE (44.2%), and incident investigation (43.2%) among the leading agentic use cases. CI/CD pipelines were identified by 36.8% of respondents, highlighting the expanding role of AI agents across software delivery and operations.
- AI agent spending grows while cost visibility lags: More than two-thirds of respondents (68.4%) report spending at least $10,000 monthly on AI agents, including 38.4% spending $50,000 or more and 12% spending at least $250,000. Yet 5.8% cannot say how much their organizations spend, while 25.8% rank inadequate cost visibility, budgeting, or spending limits among their top three challenges. Without consistent tracking of expenses by agent and workflow, organizations risk unexpected cost increases as automation expands.
Together, the findings show that engineering organizations are adopting AI agents across a growing range of operational tasks while remaining cautious about granting them unrestricted autonomy in production. As fleets expand, SRE, DevOps, and platform teams face increasing pressure to standardize agent management, maintain reliable performance, and control operating costs without adding to their own workload.
About the Survey
Komodor's 2026 Agentic Operations Survey examines the adoption of AI agents across engineering organizations, including fleet size, spending, operational use cases, governance challenges, and levels of autonomy. The research covers 1,000+ survey responses from engineering, IT, and technology roles, including cloud infrastructure, platform engineering, software development, and SRE. More than 45% of respondents hold director-level or higher positions, while nearly 85% work in organizations with at least 100 engineers, including 28.5% with 500 or more.
About Komodor
Komodor is the company behind the Komodor Agentic Operations Platform. Enterprises rely on Komodor to automate SRE, Software Operations, and cost optimization in production, and to build, run, govern, and optimize their own custom agents on the same infrastructure. Trusted by Fortune 500 companies across financial services, healthcare, retail, and more, Komodor eliminates operational complexity while improving application performance and resilience. The company has raised $90M in venture funding from leading investors in the US and EMEA. For more information, visit komodor.com, and follow us on LinkedIn and X.
Media Contact
Marc Gendron
Marc Gendron PR for Komodor
marc@mgpr.net
617-877-7480
A photo accompanying this announcement is available at https://www.globenewswire.com/NewsRoom/AttachmentNg/8c8e20d0-2db7-4cbd-bbfa-c5b88df376fa
