The promise of agentic AI is huge. But how is it impacting the enterprise and the developers and IT professionals most likely to be working with it right now? To find out, Kong evaluated labor market data and surveyed 550 tech leaders, developers, IT decision-makers, and Kong users.
In this post, we'll look at five highlights from this agentic AI adoption report, including the state of agentic AI adoption in the enterprise today, the motivation and barriers to adoption, and the split between employee sentiment and market data about what the next era of AI may mean for the tech workforce.
Want to get the full report? Download Agentic AI in the Enterprise: Paving the Path to Production.
Workers in our survey were fairly evenly split on their outlook on agentic AI’s impact on employment: 49% say AI agents will primarily support existing roles, and 45% foresee them replacing IT positions.
But market data may paint a different, more sobering picture. Employment for IT professionals and developers grew nearly 60% between 2016 and 2023. But since the rise of AI tools, job growth has stalled, increasing by less than 1% over the past two years. This period marks the weakest IT labor market since the 2010 recession.
The slowdown has also hit wages. Between 2016 and 2023, IT salaries grew roughly 50%. Over the past two years, wage growth has dropped to just 3% per year.
This uncertainty underscores the need to evolve skill sets toward higher-value work, such as designing governance frameworks, orchestrating complex workflows, and building the infrastructure that enables agentic AI. For enterprises, it reinforces the urgency of preparing systems, APIs, and governance models to handle autonomous operations at scale.
90% of survey respondents with visibility into their organization’s plans say their companies are actively adopting AI agents. This isn’t a cautious “wait and see” approach; organizations are putting strategies in place now. More than half (52%) even report having a clearly defined roadmap for adoption.
The timelines for enterprise-wide implementation are aggressive: 79% believe their organizations will reach full-scale adoption within three years.
Much like cloud migration a decade ago, enterprises are recognizing that the question isn’t whether they will embrace agentic AI, but how quickly they can do it without being left behind.
The motivations behind this rapid adoption are both pragmatic and strategic. Reducing costs and increasing efficiency emerged as the single biggest driver, cited by 23% of survey respondents. Close behind were automating IT operations (21%) and improving development and API management (20%).
When asked which teams were driving adoption, respondents pointed most often to software development (34%). Developers are at the heart of building and scaling applications, and agentic AI offers clear productivity gains by automating code testing, deployment, and integration tasks.
Other functions making big strides include security and compliance (22%), infrastructure and cloud teams (19%), and data science and analytics (16%). The fact that security and compliance appear so high on this list is worth noting: it shows that adoption isn’t just about speeding up delivery pipelines, but also about strengthening guardrails and enabling organizations to manage risk more effectively.
So where are organizations actually seeing value today? The report highlights several use cases that are moving beyond hype into measurable business impact.
Despite strong momentum, enterprises aren’t adopting agentic AI without friction. Integration complexity with existing systems emerged as the most significant barrier, cited by 31% of respondents. Retrofitting new agentic systems into fragmented legacy infrastructure is proving challenging.
A critical but often underappreciated aspect of this integration challenge is the state of APIs inside most organizations. Agentic workflows depend on agents being able to discover, access, and orchestrate APIs. But if APIs are scattered across systems without a central point of management, discovery becomes nearly impossible. Without a unified platform to bring all APIs together, the promise of agents executing end-to-end workflows breaks down.
Security and compliance concerns were close behind at 28%. As AI agents gain access to sensitive systems and data, organizations are rightfully cautious. Compliance with privacy regulations, auditability, and ensuring proper authentication are top of mind.
Notably, very few organizations doubted the business value of agentic AI. Only 12% of respondents said unclear ROI was their biggest concern, suggesting that most enterprises already see a clear path to value if they can overcome integration and governance challenges.
To tackle these hurdles, enterprises are leaning heavily on governance. Nearly three-quarters (72%) of surveyed organizations have an AI governance solution in place. Among those, over half rely on an AI gateway, a specialized middleware layer that manages, secures, and monitors interactions between AI systems and enterprise infrastructure.
An AI gateway provides policy enforcement, access control, and observability, ensuring that agentic AI can be deployed at scale without undermining compliance or introducing new risks. For many organizations, it’s the missing link that allows them to balance speed and control.
Ownership of governance responsibilities is still evolving. In many cases, platform teams (41%) are leading the charge, given their experience managing APIs and infrastructure. Others are creating dedicated AI teams (30%), while some are taking a decentralized approach, asking developers to follow guidelines without centralized oversight.
The message is clear: agentic AI is on a fast track to becoming a core enterprise function. Adoption is already mainstream, driven by practical goals like cost reduction and efficiency, and reinforced by real-world use cases such as customer support and automation.
At the same time, barriers remain particularly around integration, security, and compliance. Organizations are responding by embracing governance, with solutions like the Kong AI Gateway emerging as critical enablers for scaling responsibly.
Perhaps most importantly, agentic AI is reshaping the workforce in ways that are still playing out. While many see it as a supportive tool, labor market trends suggest that disruption is underway and may accelerate as systems become more autonomous.
The enterprises that succeed in this new era will be those that not only adopt AI quickly but also prepare their infrastructure, governance, and workforce for what comes next.
For a deeper dive into adoption trends, workforce impacts, and the strategies enterprises are using to scale responsibly, check out the Agentic AI in the Enterprise report.
Open banking initiatives have taken flight in many economies across the globe. Predicated on the open access of banking data for the overall benefit of customer choice, Open Banking comes with many challenges — security not the least of them. Givin
In Part 1 of this blog series , I broke down the "two generals problem" and shared how it affects IT leaders today. In this post, I will share the details of a specific battle from more modern history that exemplifies this concept. "Yeah, this clou
The Two Generals' Problem is a well-known thought experiment about how asynchronous - and potentially unreliable - communications can cause, shall we say, issues. [iframe src="" data-src="https://www.youtube.com/embed//xFMUFq8kC98" frameborder="0
In this blog post, we'll talk about the significant challenge of managing and governing a growing number of APIs across multiple teams in an organization — and how Control Plane Groups are a clear solution to avoid the chaos of inconsistent policies
While many organizations mistakenly believe a single tool can solve all their API security woes, the truth is far more complex. This blog post will dismantle the myth of the "silver bullet" and demonstrate how a comprehensive, defense-in-depth strat
Ever feel like you're fighting an uphill battle with your API strategy? You're building APIs faster than ever, but somehow everything feels harder. Wasn’t API-first supposed to make all this easier? Well, you're not alone. And now industry analys