TLDR: A new study by Kong Inc. indicates that 90% of enterprises are actively adopting AI agents, with 79% expecting full-scale implementation within three years. Customer support is currently the leading application, but integration complexity and security concerns remain key barriers, prompting 89% of organizations to implement or seek AI governance solutions.
SAN FRANCISCO – September 17, 2025 – A groundbreaking report released today by Kong Inc., a prominent developer in cloud API and AI technologies, highlights a significant shift in enterprise artificial intelligence adoption. The study, titled ‘Agentic AI in the Enterprise: Paving the Path to Production,’ reveals that a staggering 90% of surveyed enterprises are actively integrating AI agents into their operations. This marks a critical inflection point, moving agentic AI rapidly from theoretical discussions to practical, real-world applications.
The research underscores the ambitious timelines businesses are setting for this transformative technology, with 79% of respondents anticipating full-scale adoption of AI agents within the next three years. The primary motivations driving this rapid integration include reducing operational costs, streamlining workflows, and accelerating automation across various business functions.
Currently, customer support stands out as the most impactful use case for generative AI (GenAI), with 58% of enterprises reporting that AI-powered support systems are delivering the most meaningful business impact to date. This demonstrates the immediate value organizations are finding in leveraging AI to enhance customer interactions and service efficiency.
However, the path to widespread AI agent adoption is not without its challenges. The study identifies AI integration complexity as the top barrier, cited by 31% of respondents. Following closely are security and compliance concerns, which worry 28% of enterprises. Recognizing these hurdles, a significant majority (89%) of organizations are proactively addressing these issues by either having AI governance frameworks already in place or actively seeking robust solutions. Among those with governance frameworks, 54% are utilizing an AI gateway, an emerging ‘middleware’ control plane designed to manage AI traffic, enforce policies, and ensure compliance.
Augusto Marietti, CEO and Co-Founder of Kong Inc., emphasized the critical role of API management in this evolving landscape. “This research shows how quickly agentic AI is moving from theory to practice, but it’s also clear that adoption could stall without strong API management. APIs are critical to agents, as they can’t work without secure data connectivity or the necessary governance that goes with this,” Marietti stated. He added, “A unified platform like Konnect solves this by managing APIs in one place, provisioning credentials automatically, and making them agent-ready.”
Also Read:
- Solo.io Unveils Kagent Enterprise to Bridge AI Agent Deployment Gaps in Cloud-Native Environments
- Autonomous AI Agents Drive Workforce Transformation, Demanding Evolved Human Capabilities
The insights for this comprehensive report were gathered from a commissioned survey of 550 Kong users, developers, engineers, and IT decision-makers, providing a detailed look into the opportunities and limitations surrounding agentic AI in the enterprise.
