Enterprises are poised to invest $1.5 trillion in AI by end-2025, but true ROI isn’t coming from technology alone.
What’s being deployed isn’t just about tools, but the workforce and structure to make them meaningful. The core disconnect: AI adoption is increasingly bottom-up — frontline employees driving change — but organizations aren’t structurally ready from the top.
The paradox, then, is this: agentic AI — a new generation of intelligent systems that can reason, plan, and take action on their own — is creating a new class of employee-led innovation. The same momentum also challenges organizations to strengthen their AI governance, AI strategy, and culture, turning early experimentation into lasting advantage.
In Moveworks latest report, The New Face of AI Leadership, based on a survey of 200 IT executives, we found that the traditional IT approach—simply pushing new technology down to employees—no longer works.
In fact, more than 70% of executives believe their company underestimates the structural and cultural changes agentic AI requires.
This post gives you a guide through the three key paradoxes of this era offering a framework for a cohesive AI strategy that works.
Paradox 1: The authority shift (decentralized innovation vs. centralized governance)
Many impactful AI ideas now originate beyond IT. They’re coming from frontline employees redesigning their own workflows.
The evidence:
- 91% of executives say non-technical employees are playing a larger role in AI-agent projects.
- 78% of executives are in favor of letting good AI-agent efforts come from anywhere in the company.
- Only 38% of execs view IT as the top AI-influencer over the next three years.
This creates a leadership vacuum. As AI control becomes more dispersed, governance, risk control, security, and alignment mechanisms need to be set up to enable distributed innovation.
If not, organizations risk disorganization, duplication, data risk, or strategic drift. Leaders can adapt by shifting from “AI Owner” to “AI Platform Enabler,” to empower bottom-up energy, while keeping AI strategy, risk, and scale in view.
How to help manage:
- Charter an AI governance and implementation committee (cross-functional: IT, HR, operations, risk).
- Build a service catalog/AI agent marketplace. Non-technical teams can launch ideas within pre-approved guardrails.
- Define data, security and usage policies in advance. Make them accessible and simple.
Paradox 2: The new AI career path (technical skill vs. operational expertise)
The “AI talent gap” isn’t just about hiring more data scientists. It’s about empowering the existing workforce to become “AI-fluent” in their operational domains. Wages are rising twice as fast in AI-exposed industries, even in roles previously considered highly automatable.
AI agents that understand user goals, make decisions, and take action across applications —are at the center of this transformation.
These autonomous systems act as hands-on building tools: helping non-technical employees experiment, automate, and apply AI directly within their day-to-day workflows. Instead of learning AI in theory, workers gain familiarity by collaborating with AI agents that reason, plan, and execute tasks and support countless new use cases alongside them.
A Finance employee might use an AI agent to process high-volume purchase orders by verifying vendor invoices against budget allocations, while a Human Resources employee utilizes an AI agent to screen hundreds of resumes and shortlist candidates based on job requirements.
The evidence:
- 39% of executives expect agentic AI to create upward mobility for all employees, not just technical specialists.
- Nearly 75% of employers struggle to find skilled talent. Only 16% confident in their tech teams.
- Global AI talent gap widened in 2024-25 due to explosive demand and insufficient supply.
- Almost 70% of companies have already created new roles (like AI project coordinators, content reviewers) to manage agentic AI adoption.
- The hiring split: 54% of leaders prioritize technical AI expertise, 47% prioritize strong ideas for applying AI to business challenges.
Leaders are often split between hiring deep (but costly) technical talent such as data scientists, ML engineers, and others, or hiring automation and operational experts who understand business workflows and can apply AI to real world scenarios.
Hiring more AI specialists isn't the only way to find more AI talent. Sometimes, empowering your existing workforce with the right AI tools can be. Consider how Coca-Cola Consolidated empowered business users to collaborate with IT during an internal hackathon, an event where they built dozens of new use cases, highlighting how bottom-up creativity and top-down structure can move in sync to accelerate transformation.
If you focus only on bringing on AI talent without considering your existing people and culture, you can miss countless opportunities visible only to the employees and operational experts closest to them. Yet nor should you give employees unfettered, unchecked access to AI tools. You should chart a middle ground, opening the door for teams to safely and securely experiment with AI with clear guiderails and policies in place.
How to help manage:
- Source ideas from every corner of the organization. Some of the most creative AI use cases come from non-technical teams closest to the work.
- Identify emerging AI champions across departments, from marketing and finance to supply chain and HR, and invite them to shape your AI roadmap.
- Create “AI fluency” training tracks that blend practical use cases, workflow design, governance, and responsible experimentation.
- Launch new role definitions such as AI Project Coordinator in operations, AI Content Reviewer in communications and HR, Domain-Agent Owner within business units that are experimenting with agentic AI.
- Reward exploration and sharing. Highlight individuals and teams who use agentic AI to solve real problems, and make their success stories part of your growth culture.
Read the Report: Moveworks Report Reveals Employee-Led AI Is Redefining How Enterprises Work
Paradox 3: The adoption divide between radical reimagination and existing integration
Leadership knows that true transformation often requires fundamentally rethinking workflows, with 53% citing process reinvention as ROI. Yet, employees often prefer tools that integrate with their current processes rather than re-imagine them. Think of an AI agent that surfaces Jira tickets or updates a Workday record right inside Microsoft Teams — there’s no new app to learn, no extra steps required either.
More than 60% of executives say their employees prefer AI tools that integrate into existing workflows rather than completely reimagine them.
That finding underscores the tension and opportunity of agentic AI: its ability to both reinvent workflows and seamlessly fit into the systems people already use. AI agents help strike a critical balance by enabling huge transformation, but also supporting short term gains with immediate incremental changes.
The evidence:
- 73% of executives believe their company underestimates the structural and cultural changes required by agentic AI.
- 65% of executives say their employees prefer AI tools that integrate into current processes rather than ones that demand re-thinking everything.
So how can you get the best of both worlds: game-changing enterprise transformation when your people are asking for incremental, low-friction improvements?
Organizations that adopt radical change too quickly may experience resistance, confusion, and backlash. If you go too incremental, you risk staying behind.
Ultimately, it’s a mix of balancing quick wins by integrating AI to solve immediate, visible pain points (password resets, approvals, content generation), and also publicly showcasing high-impact projects where AI has completely re-imagined a major workstream. Lead from the top, share stories, build aspiration.
How to help manage:
- Consider launching a “first-wave” of agentic AI pilots in low-risk, high-visibility areas (e.g., IT support, HR onboarding)
- Communicate: share “why we’re doing this” in employee-facing language. Explain the integration vs reimagination tension
- Highlight “moonshot” projects where a major workflow has been reinvented. Use them as storytelling tools
- Embed feedback loops: Measure and track success metrics, and ask employees what’s working, listen, and iterate
By balancing incremental trust-building with long-view vision, you keep the organisation engaged and aligned.
Your action plan for building an effective AI strategy
The AI talent and culture paradox is the central challenge for modern leaders. You cannot control the full transformation, but you can guide it. You can balance top-down strategy with bottom-up empowerment.
The path forward is less about selecting the right tool, and more about enabling talent, culture, and structure to scale innovation safely and strategically.
Key actionable steps:
- Elevate governance: Form an AI implementation committee to act as strategic partner, not gatekeeper and train employees on secure usage.
- Rally and upskill champions: Formally identify, reward, and share success stories from your non-technical innovators.
- Communicate with purpose: Acknowledge the “integration vs reimagination” tension; create tailored messages for both audiences.
- Refine through feedback: Create formal feedback loops to monitor metrics, optimize workflows, and show employees you are listening.
The leaders who succeed will be those who stop managing technology and start enabling talent. Those who build a culture ready for the agentic AI era, not just the next tool.
When you lead with people, governance, culture and vision in equal measure, you turn the paradox into a competitive advantage.
Read the Report: Moveworks Report Reveals Employee-Led AI Is Redefining How Enterprises Work
Table of contents
