Why AI Adoption Is Really About People, Not Tech

AI is everywhere, yet many organizations still struggle to motivate people to actually use it. The problem isn’t the technology, it’s human behavior. Understanding these human barriers is the difference between AI pilots that stall and adoption that can truly scale.

A recent Harvard Business Review article, Why People Resist Embracing AI, makes the case clearly: AI succeeds or fails based on whether people trust it, understand it, and believe it supports rather than replaces them. In our work as consultants and change management experts, we’ve come to understand these common barriers to adoption and how to mitigate them.

Additionally, over the last year, we ourselves have lived through the process bringing AI into our workspace and culture. Here are a few common points of resistance and some solutions that we have translated from our own lessons learned.

Concern # 1: AI Feels Like a “Black Box”

Many AI tools make decisions without clearly explaining the why behind the outputsor what logic they’re based on. When outcomes feel mysterious or unexpected, trust drops quickly for the user.

Solution:

When interacting with Generative AI tools (e.g., Microsoft Copilot, ChatGPT, Claude), start simple. As a simple fix, start by asking AI to explain its reasoning. Include text in your prompt that asks how it’s getting to the information it provides. That may help you in the process of reviewing and validating outputs, ensuring it doesn’t feel too obscure.

How FMP turns this into action:

• Established Momentum AI as a unifying internal initiative to align adoption, governance, and workforce capability building around a shared strategy. This approach makes AI feel less opaque and enabled knowledge-sharing across our workforce.
• Delivered targeted communications and practical resources (e.g., tip sheets, guides) to coach employees on how to build more effective prompts and validate outputs.
• Developed and delivered varied, low-barrier learning formats, such as including web-based training, to meet employees where they are and drive more confident adoption.

Concern # 2: AI Feels Too Cold for Human Work

The aforementioned HBR article also states that people are generally comfortable using AI for objective tasks like data analysis, but far less comfortable using it for judgement-heavy or creative work.

Solution:

To address this, you can frame AI as support, not a replacement. When AI handles the more objective parts of work, people can focus on what requires human insight.

How FMP turns this into action:

• Used an employee spotlight series, Momentum Moment, to normalize AI as a teammate, not a replacement, and share real-world successes and challenges.
• Produced two short videos and several employee-led video podcasts to showcase real use cases and step-by-step application.

Concern # 3: AI Seems Rigid or One-Size-Fits-All

Many of us have had the same experience: we write a prompt for a GenAI tool and receive an output that seems generic, off base, or unusable. That moment often reinforces the perception that AI is rigid or one-size-fits-all and that using it effectively requires extra effort just to get to a workable result. When first starting out with AI, it’s easy to miss that the tool’s performance is highly dependent on context, feedback, and iteration.

Solution:

With clearer inputs and refinement, outputs improve dramatically. As people practice (e.g., learning how to add context, adjust prompts, and respond to what the tool produces), they gain more control than they expect. That experimentation leads to better results, and those better results are what makes AI feel usable, scalable, and genuinely relevant rather than generic.

For example, someone might ask AI to “draft an email announcing a new process” and get back a bland, overly formal message that doesn’t fit their audience. That result can feel like proof that AI only produces cookie-cutter content. But when the same prompt is refined (adding details about who the audience is, what concerns they’ve raised, and the tone that’s needed), the output changes significantly. With that added context, the message becomes more specific, more human, and far more usable. Less rework is needed too.

How FMP turns this into action:

• Designed internal initiatives around experimentation and iteration (e.g., peer-led workshops, office hours), encouraging learning over perfection.
• Developed detailed instructor-led training on prompt engineering and context engineering to improve both prompts and outputs, pairing training with supplemental resources (e.g., tip sheets, reference guides) to reinforce learning.
• Launched an AI super user group to share prompts, lessons learned, and real examples that are flexible and customizable.

Concern # 4: AI Feels Like It Takes Away Control

When AI operates too independently, people worry about losing decision-making power or influence over the outcome. However, even small opportunities for human input (e.g., reviewing, adjusting, or guiding outputs) go a long way to increase comfort and trust.

Solution:

Build in human interaction all along your workflow when using AI-based tools. Whether it’s reviewing work or leveraging others to review, test, and adjust, make sure you’re continuing to shape the work as the human in the loop.

In practice, this can look as simple as building review points into an AI‑supported workflow (i.e., draft using AI first, have a human review, use AI to refine). Each touchpoint reinforces that AI is assisting the work, while people continue to shape and own the final result.

How FMP turns this into action:

• Increased trust by clearly defining guardrails and establishing internal policies, so employees understand where AI fits and where human judgment remains essential.
• Reinforced shared accountability through a supporting governance framework, change management plan, and evaluation approach, alongside targeted training.

Concern # 5: People Still Prefer People

Even when AI performs as well as humans, many people still prefer human interaction. This is especially true when it comes to meaningful or sensitive work. As organizations implement AI and revisit workflows, decisions are being made around what work can be done through AI-based tools and what work is best reserved for humans.

Solution:

It’s often best to design AI-based tools, deliverables, and applications that complement human interaction rather than replace it. AI should serve as a teammate, not a takeover. Transformation doesn’t need to be all‑or‑nothing; it can start simply and grow from there.

How FMP turns this into action:

• Built a foundational AI maturity model with shared language and measurable markers to help teams assess adoption, scale AI, and maintain human collaboration (and decision-making) across workstreams.
• Anchored AI adoption in people-first change practices, including executive-led fireside chats, supervisor-focused training, and monthly collaboration across workstreams to create space for dialogue, questions, and shared learning.

Looking back on 2025, we see a year defined not solely by technological change, but by collective curiosity. The question was never just how quickly we could adopt AI, but how thoughtfully we could bring our people along on the journey. Through a variety of channels (Figure 1), we approached AI adoption as a change journey, not just a technology rollout. The organizations that succeed will invest as much in trust, clarity, and learning as they do in tools.

At FMP, we center AI strategies around human psychology and organizational readiness so that adoption is intentional, responsible, and scalable. When people understand where AI fits and where human judgment remains essential, adoption follows naturally.

Read more about how AI can transform your organization and FMP can help:

• “AI in the Workplace: A Peek into FMP’s Policy”.
• “From Mandates to Momentum: A Human-Centered Approach for AI Adoption in Government”
• “Navigating AI Adoption with Confidence: FMP’s AI Maturity Model”
• “Unlocking the Power of AI through Workforce Fluency”
• “Building Momentum: How FMP Uses AI to Deliver More Innovation Solutions”               
• “Sustainable AI: Using Innovation Responsibly”

Jessica Waymouth joined FMP in 2014. She is a Managing Consultant helping organizations drive lasting change by aligning people, strategy, and systems. She co-leads FMP’s Strategic Communication Community of Practice (CoP) and brings a thoughtful, results-driven approach to organizational transformation. She has a particular passion for mission-driven impact, designing environments that empower individuals and organizations to grow. Outside of work, she’s a mom of two, curious traveler, local volunteer, and loves a good book.  

Emma Wright joined FMP in 2020. She is a Senior Consultant in the Learning and Development Center of Excellence and is the FMP Blog Editor. She supports a variety of initiatives across multiple clients, including program management, strategic planning and communications, and training and development. She hails from Alexandria, Virginia, and you can often find her cooking, out at a concert, or eating at her favorite DC restaurants.