Artificial intelligence is no longer a futuristic concept; it’s a present-day reality for most IT professionals. A recent global study by Alteryx, surveying 700 data analysts and 700 IT leaders, reveals a striking trend: a staggering 96% of IT professionals are already leveraging AI in their daily work. This widespread adoption signals a significant shift in how organizations approach data and decision-making.
However, while adoption is high, the depth of engagement varies. Only about half of these professionals, 49% to be exact, consider themselves frequent users, integrating AI tools “always or most of the time.” This suggests a landscape where AI is present, but its full potential is still being explored and integrated into core workflows.
The Rise of Agentic AI and Future Outlook
The conversation around AI is rapidly evolving, with “agentic AI” emerging as a key area of focus. These autonomous AI agents can perform tasks, make decisions, and interact with systems with minimal human intervention. The Alteryx study found that nearly six in 10 respondents, 59%, anticipate actively employing AI agents within the next 12 months, highlighting a strong appetite for more sophisticated AI capabilities.
A notable finding is the willingness of professionals to grant these agents significant autonomy. Roughly half of the respondents expressed readiness to provide AI agents with “unrestricted access” to their data. While the survey report didn’t delve into the security implications, it’s clear that human oversight remains a critical concern, with 44% of professionals emphasizing its necessity alongside such access.
Currently, the most common applications for agentic AI include practical tasks like drafting communications and streamlining scheduling workflows. These initial deployments showcase AI’s immediate value in automating routine processes and freeing up valuable human time.
The Hidden “AI Tax” on Productivity
Despite AI’s promise, the study uncovers a significant “AI tax” on professionals’ time, particularly in data preparation and validation. Data analysts report spending close to six hours per week on “foundational data work”—cleaning and prepping data for AI models and retrieval-augmented generation platforms. A substantial 48% even dedicate six to ten hours weekly to these tasks.
Interestingly, the tools of choice for this foundational work haven’t entirely shifted to cutting-edge AI platforms. Spreadsheets still dominate, cited by 61% of respondents, followed by business intelligence tools (56%) and dedicated data preparation platforms (51%). This trend suggests that AI is often “layering on top of existing workflows rather than replacing them,” as the report authors note, meaning traditional tools remain deeply embedded.
Beyond preparation, validating and correcting AI-generated outputs adds another layer of time commitment. Analysts spend nearly four hours per week ensuring AI results are accurate and reliable, with one in six spending six hours or more. When combined, these foundational tasks and validation efforts amount to almost ten hours per week—nearly two full workdays—an AI-driven overhead that organizations must address.
The Critical Need for Human Oversight and New Skills
The journey from data analysis to business decision-making is also not as swift as many might assume. Despite the focus on real-time capabilities, only 20% of organizations can move from analysis to decision within a few hours, and a mere 5% support genuine real-time decision-making. This gap highlights a significant area for improvement in operationalizing AI insights.
Key barriers to integrating AI into business decisions include the challenge of explaining complex AI outputs to decision-makers and a noticeable lack of analytical skills across businesses. This underscores the need for better communication and training to bridge the understanding gap between AI capabilities and business strategy.
Ultimately, the study points to an emerging and increasingly valuable skill set in the AI age: validating AI outputs. While AI can undoubtedly accelerate work, the findings emphasize that “organizations still need human oversight to ensure outcomes are consistent, explainable, and trusted.” This highlights that the future of AI isn’t about complete automation, but rather intelligent collaboration between advanced technology and skilled human insight.
Source: ZDNet – AI
