As the world races toward autonomous AI solutions, recent data from Gartner has sparked industry-wide introspection. The firm forecasts that over 40% of Agentic AI projects will be canceled by the end of 2027, citing escalating costs, unclear business value, and inadequate risk controls as primary reasons. In a January 2025 Gartner poll of 3,412 enterprise webinar attendees, only 19% reported significant investments in AI agents, while 42% took a conservative stance, 8% abstained altogether, and 31% remain undecided or cautious.
In response to these sobering projections, CP Gurnani, Co-founder and CEO of AIonOS, offered a candid perspective: “Frankly, I think Gartner is being conservative when they say 40% of super-agentic AI experiments will get canceled. In my opinion, it could be closer to 50%. However, it’s not necessarily a failure of the technology. It’s about the interface journey and adoption readiness. Version 0 or Version 1 of any product is always the hardest to adopt.”
Gurnani likened the current wave of Agentic AI adoption to earlier tech evolutions: “People who remember the early days of MS Word or spell-check tools will know what I mean. Back then, failure rates were high simply because the tools were not intuitive. Agentic AI is facing the same curve.”
Gurnani emphasized that these initial setbacks shouldn’t be seen as the downfall of Agentic AI. Instead, they represent the growing pains of a transformative technology. “No super-agentic AI can thrive without constant learning, data feedback loops, and iterative improvement. Abandonment does not equate to failure but it’s part of the maturation cycle.”
“Failures will be visible and loud,” said Gurnani, “but the underlying technology will prevail. Just like early tools that stumbled before becoming indispensable, Agentic AI will become smarter, faster, and more intuitive. The winners will be those who stay invested in solving real problems, not chasing trends.”
Focus on Enterprise Value, Not Hype When it Comes to Agentic AI
Arjun Nagulapally, CTO of AIonOS, reinforced this sentiment by underlining the importance of building human-in-the-loop systems that deliver real-world value to clients. “Agentic AI, as we define it, is about enabling autonomous activities with humans in the loop. We don’t believe AI should fully replace human decision-making. Instead, it should eliminate drudgery and elevate strategic focus.”
He added that AIonOS is not chasing hype cycles or proof-of-concepts: “We’ve been deploying agentic solutions with actual enterprise customers for nearly a decade. These are real deployments yielding measurable results. We are not experimenting but we are executing.”
Nagulapally also addressed skepticism about full autonomy in AI systems: “The technology is not fully there yet for end-to-end automation. High costs and complexity remain limiting factors. But that does not diminish the real value that well-architected, human-assisted AI can bring today.”
As the AI landscape matures, companies like AIonOS are betting on incremental progress, human-machine collaboration, and enterprise-focused delivery to outlast the hype and build sustainable success in the age of Agentic AI.
(These views were shared along the sidelines of the event where AIonOS announced its acquisition of a majority stake in Cloud Analogy)
