How enterprises will deploy agentic AI and autonomous BI: - The biggest change I expect in 2026 is seeing agentic AI shift from pilots to production, taking on multi-step (very prescribed) workflows across non-IT aspects of their business such as operations, finance, HR, and customer support. (As an AI solution implementer, I see that agentic adoption in production is actually pretty low in 2025 (see study that backs what I see, stating only 33% of firms actually deploying agents to production, https://kpmg.com/us/en/media/news/q2-ai-pulse-2025-agents-move-beyond-experimentation.html, but I expect it to increase in 2026) - Enterprises will finally "get" the power of agent/LLM reasoning, planning, and taking action across systems without humans directly involved. - BI/analytics will shift from basic dashboards to continuous insight generation, where agents watch data streams and surface insights, issues, or opportunities in near real time. - CIOs will feel pressure from business leaders to move beyond dashboard upgrades and focus on helping the business automate real decisions, (using agents) that will turn insights into action without needing a human in the process. This is interesting because it will dramatically change how business is conducted, will have a massive impact on productivity, and is also something most businesses are not yet doing and it requires the most complex part of AI (multi-agent workflows) to be developed. What the shift from departmental RPA to enterprise-wide orchestration and outcome-based pricing means for CIOs: - Traditional RPA never lived up to its promise. The original technology was too brittle and didn't deliver the hype it promised. But, modern RPA (also referred to as AI Automation) is evolving and integrating enterprise-class automation platforms that coordinate agents, APIs, and workflows across the entire business ecosystem rather than isolated tasks inside departments. - CIOs will want to select and deploy a company-wide AI platform and better control what AI tools are being used to keep everything secure and well-governed. - I see the most successful/impactful AI automation (modern RPA) from companies that are willing to rethink an entire process and not just try to wedge AI into nooks and crannies of an existing process. This means automation initiatives should spread across the whole organization, not only single departments, and this requires strong change management to be adopted and to bring ROI.
In the coming year, businesses will transition from testing agentic AI and autonomous BI as pilots to running them as operators by embedding these tools into existing ERP, Finance, and Customer Experience processes. Instead of using the tools to create reports on historical performance, businesses will use the tools to continuously monitor the status of their operations, propose corrective action, and implement approved changes involving low-risk transactions. Essentially, the change will transition AI from being a source of information about business activities to becoming a source of guidance for specific elements of the business. The shift from departmental automation to enterprise-wide orchestration particularly with outcome-based pricing requires CIOs to think like owners of their products and profit-and-loss statements. Rather than measuring success in terms of the quantity of automated transactions, success is now measured by the total financial benefit returned, cycle times reduced, and compliance risks eliminated, achieved through coordinated enterprise governance, new key performance indicators, and a unified automation architectural framework that operates as part of a company's internal cloud services. Therefore, establishing AI-ready Data Governance will become the key enabler for businesses to use agentic AI and form the foundation for the development of autonomous agent-based systems that will operate in a manner that is safe and consistent in nature. Additionally, ironically enough, greater exposure to AI technologies will spur the acceleration of modernizing projects that expose clean Application Programming Interfaces, events, and real-time data, all the necessary raw materials for developing intelligent systems. In my opinion, by the year 2026, CIOs who view agentic AI as a formally-engineered discipline that employs service level objectives, rollback procedures, and measurable outcomes will have a competitive edge over those CIOs who merely use agentic AI as a demonstration of cutting-edge technology and operate the system using existing business processes.
Agentic AI and BI automation will have a big impact on organizational structures. At the moment, there is a clear divide between analysts and engineers. But when AI tools can generate SQL from natural language, build pipelines and handle most of the execution, this divide starts to collapse. Only a few, senior-level roles will likely remain, focusing more on the management of AI agents who take over the entire lifecycle of data collection and analysis. These people, which I believe are going to be the CIO/CXOs, will then make decisions to ensure that organizations operate efficiently with AI-first systems. Most readers are probably aware that this trend does not gather traction only in the business intelligence sector where we operate, but in all industries where one department relies on another for its task execution.
Agents and Automation Technologies Have Reached New Levels. CIOs Are Now Ready to Have These Types of Technologies Perform Their 1st Pass on Decision Making Versus Just Dashboarding as Before. The Shift From Deploying These Technologies as Experiments to Deploying Them as Digital Operators With Defined Accountability is Where I Am Seeing the Largest Change. Moving Away from Dispersed RPA on an Enterprise Scale to Orchestration Creates A Greater Value for Enterprises. In Every Instance That I Have Seen CIOs Combine Their Automated Processes to Become an Organization-Wide Process Has Resulted in a Greater Efficiency Because Users No Longer Continue to Redevelop the Same Process in Multiple Business Units. CIOs Must Focus on Establishing A Solid Data Governance Framework to Support AI Development Efforts in 2021 and Beyond. In My Experience, I Have Seen All Failed AI Implementations are Due to Poor Quality, Unvalidated Data. CIOs Who Treat Data Governance as A Layer of Performance and Not as Documentation Will Scale Faster. While There Is A Lot of Hype Surrounded By Technology Modernization, This Does Not Mean Looking Away From It. AI Technology Is A Double-Edged Sword in Terms of Revealing The Weakness of Legacy Systems That Have Been Ignored. If It Is Not Running on A Clean and Real-Time Foundation, AI Technologies Will Not Be Able to Deliver Optimal Value. I Would Encourage CIOs to Be Prepared For Next Year By Recognizing the Demand for Transparency Within Every System as A Result of AI Technologies. The More Rapidly They Recognize This Demand, the More Successful Their AI Technology Deployment Will Be.
Automotive, claims, and supply chain operations teams will do more than run pilot projects or expand automation learning loops. They will act and automate using agentic AI (for continuous real-time decisioning) and autonomous BI (for instant closure of long standing claims handling, repair supply chain, and customer journey gaps) to eliminate backlogs, leakage, and legacy processing. CIOs who have focused on business unit-specific robotic process automation (RPA) will be asked to scale up to organisation-wide orchestration and workflow exception handling with consumption-based pricing for results. Success will be determined by combining point solutions that automate in isolation of one another with integration solutions that focus on unifying disconnected claims, telematics, dealer-management, and other operational data. AI governance, which includes the management of MLOps and autonomous BI, will be table stakes as insurers and automotive dealers navigate incomplete, unstructured, and unusable document capture; unstructured supply-chain data; demand for regulatory and compliance explainability; and continued core systems modernisation for any autonomous decisioning loops that are introduced to automate claims processing, fraud, or supply chain risk. CIOs also will face a bigger change: a move to continuous closed-loop (zero latency) operations as core workflow triage, supply chain logistics for parts and resources, automated fraud checks, and communications with claimants and customers execute autonomously and on a continuous, real-time basis, impacting cost structures, operating cycle times, and industry differentiation.
Having consulted for multiple organizations across the travel and logistics segment, the consensus is that, in 2025, organizations will move away from testing with agentive AI and move towards utilizing it on a daily basis. At LAXcar, we utilize small agent workflows to unsupervised, goal-oriented workflows that we use, and I expect chief intelligence officers to scale this to autonomous ordinal Business Intelligence (BI), where it is, perhaps, uncategorized. This will lead to a shift from 'passive dashboards' to real-time BI that is ready for action. The shift from siloed Robotic Process Automation (RPA) to the orchestration of mobility across units will also lead to expert status, and will lead to a big increase in outcomes from automation, no longer framing improvements in hours, and the removal of manual work. I have experienced this with end-result-focused tactics from suppliers, which were implemented first, outlining the return on investment (ROI) to be targeted. None of this will proceed without AI-ready, bias-specific governance. Unified and comprehensive datasets are the baseline. I expect the so-called modernization of IT to be less prominent, not more, as we will keep legacy systems as they are and add a low-code structure on top of existing systems to keep real-time data from being compressed. The biggest trend I have witnessed is a focus on the confluence of automation, AI, and BI into one seamless operating structure.