AI Usage Polls and Studies

In this edition: an overview of three key reports and their findings from Gallup, Claude, and Harvard Business Review.

This is Q&AI, our blog series aimed at keeping you in the know on updates in the rapidly evolving world of AI. Sometimes, these will be quick updates on new developments in the field. Sometimes, they’ll be tips on tactics, features, or functionality. If you haven’t met me yet, hi: I’m Jen Taylor, CI’s Director of AI Strategy & Implementation, and your (very human) AI BFF. 

Q: What is there to learn from recent research into AI’s impact in our day-to-day lives?

A: I’m going to focus on three recently-released reports and their findings.

From Gallup: 2025 AI Usage

A new Gallup poll looking at AI usage in 2025 shows that while early 2025 saw growth in overall AI usage, that growth flattened by Q4. What continued to increase, though, was frequency. The people already using AI are using it more often. So we are seeing consolidation, not broad expansion.

What stands out most by the end of 2025 is who is using AI. Leaders and employees in remote-capable roles are using it the most. Individual contributors and non-remote roles are significantly behind. That gap is widening, not closing.

We are also still seeing low organizational ownership. Only 38% of employees say their organization has integrated AI. Even if “integrated” does not necessarily mean formal policy, it does suggest that most organizations do not have clear, visible AI strategy or enablement in place. The 21% who say they do not know reinforces that communication and alignment are weak.

One important tension point is middle management. Leaders are using AI heavily. Individual contributors are lagging. Managers sit in between. Their adoption has grown, but they are not at leadership levels. Since managers are the ones translating strategy into daily work, uneven adoption at this level likely slows broader organizational acceleration.

Uneven adoption means uneven acceleration and uneven organizational benefit. Knowledge workers and leaders are compounding productivity gains. Frontline roles and less remote-capable positions are not accelerating at the same rate. Without clearer enablement, use cases, and manager support, that gap will likely widen.

From Claude: AI Study on Coding Skill Development

Anthropic recently released a report looking at how using AI to code impacted skill development. Unsurprisingly, participants who used AI scored lower on a follow-up test than those who hand-coded. The AI group showed weaker mastery. The headline takeaway is straightforward: using AI while learning something new can reduce skill development.

The more interesting nuance is that it was not simply AI use that mattered, but how it was used. Participants who delegated heavily to AI learned the least. Those who used AI to ask conceptual questions, request explanations, or generate code and then interrogate it retained far more understanding. AI can either replace your thinking or support it. You get to choose whether you use it to produce outputs for you or to think alongside you. That interaction pattern appears to determine whether skills erode or deepen.

From Harvard Business Review: “AI Doesn’t Reduce Work—It Intensifies It”

This study validates what an audience member said to me once after I shared the SF Opera case study at a conference: AI gives you more, different work to do. In an 8-month study across a 200-person tech company, the research observed:

1. Task expansion – people started doing work outside of their official scope
2. Boundary erosion – people continued working after hours, because it was fun and didn’t feel like work
3. Multitasking overload – increased context switching and cognitive load

This identified that while AI can reduce time on some things, the incremental capabilities can make you feel more stretched.

TALK TO YOU SOON,
Jen

Have a question for a future edition? Submit it here!