Qlik®, a worldwide chief in information integration, information high quality, analytics, and synthetic intelligence (AI), in the present day launched its Qlik 2025 Agentic AI [HK1] Research, a take a look at how giant enterprises are planning, funding and operationalizing Agentic AI. The research, commissioned from Enterprise Know-how Analysis (ETR), confirmed sturdy dedication to brokers, paired with clear execution gaps. Practically each respondent has dedicated funds, but most say it can take years to operationalize at scale, with information high quality and integration of Agentic AI initiatives[HK2] with current techniques cited because the main blockers.
“Enterprises will not be brief on ambition or funding. What’s lacking are the info and analytics foundations that allow brokers work throughout the enterprise with reliability and management,” mentioned James Fisher[HK3] [EP4] , Chief Technique Officer, Qlik. “If you’d like Agentic AI to maneuver the needle in 2026, make investments first in trusted pipelines, interoperability, and a sensible ROI framework your board believes.”
Key findings
- Cash is in
97% have dedicated funds to Agentic AI, with 39% planning to spend 1 million {dollars} or extra and 34% allocating 10 to 25 p.c of their AI funds. That is now a line merchandise, which units expectations for seen ends in 2026. - Technique is maturing, however worth measurement lags[HK5]
69% report a proper AI technique, in comparison with 37% in 2024[EP6] , but solely 19% have an outlined ROI framework. The governance story is shifting from “ought to we” to “what did we get for it.” - Scale will take time
Solely 18% have absolutely deployed Agentic AI [HK7] and 46% say scale is three to 5 years away, with simply 42% assured of their inside experience. 2026 seems like a construct section, not a broad rollout. - Information is the choke level
Information high quality, availability and entry lead the barrier checklist, adopted by integration, expertise and governance. The constraint is enterprise plumbing, greater than [HK8] mannequin horsepower. - Danger sits on the deployment edge
High issues are cybersecurity, output reliability and authorized publicity, with explainability and auditability shut behind. Danger leaders will form tempo and vendor choice. - The place brokers land first
IT operations and software program improvement are probably the most focused areas, with price discount the highest aim and productiveness the important thing metric. Early wins cluster the place telemetry and baselines exist already.
Agentic AI has crossed the funds threshold and moved into 2026 working plans. The early motion is pragmatic: respondents level to IT operations and software program improvement as first movers, the place outcomes and telemetry are clearer. The drag on scale is much less attributable to mannequin functionality andextra carefully associated to the problem of integrating ruled[EP9] [KJ10] [EP11] , high-quality information into current workflows and connecting techniques with out including threat. Till that occurs, many packages will stay pilots and proofs quite than run-rate operations.
“As spend shifts from experimentation to line gadgets, the constraints are basic enterprise ones: information high quality, integration, governance and expertise,” mentioned Erik Bradley, Chief Strategist, Enterprise Know-how Analysis (ETR). “Our information reveals broad intent, however solely a minority are able to scale. The following yr will probably be about turning tightly scoped use instances in IT ops and software program engineering into sturdy, measured manufacturing.”