Your Top AI Adoption Questions Answered: Webinar Recap

The rush to implement AI is real, but the path to successful adoption remains unclear for many organizations. In Prosci's recent global webinar series, over 10,000 change management professionals registered to explore a critical question: How do we move beyond AI deployment to genuine AI adoption?

The conversation between Tim Creasey, Chief Innovation Officer and Paul Gonzalez, Senior Director of Product, revealed critical gaps between AI's technical potential and organizational reality, offering practical guidance for change practitioners navigating this transformation.

The Leadership Foundation Challenge

Twenty-five years of Prosci research consistently shows active and visible leadership participation as the top contributor to change success. AI adoption amplifies this need in unique ways. Leaders must move beyond announcing AI initiatives to personally testing and understanding these tools' boundaries and opportunities.

Prosci research reveals a critical distinction between organizations with smooth AI implementations versus those struggling. Successful implementations demonstrate strong leadership support, while struggling organizations show weak or inconsistent executive engagement. This isn't just about sending emails or holding town halls: it requires leaders to model the AI behaviors they expect to see.

Paul's work with change practitioners has identified leading indicators of AI adoption failure. When CEOs approach AI from a position of fear—scrambling because competitors are implementing it—they create reactive rather than strategic initiatives. Similarly, when executives announce AI adoption then immediately delegate it to IT departments, they signal this is a technical implementation rather than a fundamental transformation of how work happens.

The most successful AI implementations combine what Paul calls "bold AI vision" with sustained executive involvement. Leaders need to articulate not just what AI tools will be deployed, but how the organization will fundamentally change as a result.

The Sanctioned vs. Unsanctioned AI Paradox

One of the most intriguing challenges facing organizations today is the gap between personal and enterprise AI adoption. Poll results from the webinar showed individual AI adoption consistently outpacing organizational enablement: a phenomenon rarely seen in traditional technology implementations.

Unlike previous technology rollouts where employees waited for organizational deployment, AI tools are accessible through personal browsers and applications. Well-intentioned employees are racing ahead, seeking ways to improve their work performance and customer outcomes. However, many operate in policy gray areas, uncertain about what's permissible.

This creates both opportunity and risk. The employees advancing fastest with AI often discover the most valuable use cases and develop proficiency that could benefit entire teams. Yet they may also be working in isolation, creating performance disparities and potentially violating unclear organizational policies.

Change practitioners need to find ways to illuminate this unsanctioned AI use positively, encouraging high performers to share their discoveries while establishing clear guidelines for acceptable use. Organizations must also address the quality gap between consumer AI tools and enterprise solutions, providing transparency about capabilities and roadmaps for improvement.

The Training Half-Life Problem

Prosci's research identifies insufficient training as nearly 40% of AI adoption challenges. However, traditional training solutions face a unique obstacle: the half-life of AI skills is approximately three to four months. Features change, capabilities expand, and new tools emerge so rapidly that specific training becomes obsolete quickly.

This creates what Paul calls a "training paradox." People need more training, but investing in specific tool training yields diminishing returns as capabilities evolve. Instead, organizations need to focus on two critical areas.

First, building continuous change capability throughout the organization. AI adoption isn't a project with a defined end state; it's an ongoing transformation requiring sustained change management skills. Leaders need support fulfilling their CLARC roles (Communicator, Liaison, Advocate, Resistance Manager, Coach), sponsors need to maintain the ABCs of sponsorship, and employees need tools for navigating continuous ADKAR journeys.

Second, developing what Tim calls "AI literacy"—not technical skills, but pattern recognition for when AI can enhance work. The AI Integration Framework helps individuals categorize their tasks into three buckets:

• My work: Human-exclusive tasks requiring emotional intelligence, ethical judgment, or complex interpersonal interaction
• With me work: Collaborative tasks where AI can enhance human capabilities, like drafting first versions, accessing expertise, or providing alternative perspectives
• For me work: Routine, repetitive tasks that AI could potentially automate completely

This framework provides a stable mental model even as specific tools and capabilities evolve. Rather than learning button clicks and feature locations, people learn to ask: "When might I bring a digital collaborator to help me achieve better results faster, with higher quality, less mental strain, and more enjoyment?"

Moving Beyond AI Adoption to AI Proficiency

Change practitioners risk over-indexing on speed of adoption when measuring AI success. Simply logging into an AI tool and entering a prompt may count as "adoption," but the real value lies in sustained utilization and proficient integration into daily workflows.

Prosci's change performance framework provides a more comprehensive measurement approach, tracking not just initial usage but retention rates, depth of engagement, and skill development over time. Organizations need visibility into who their power users are, what use cases drive the most value, and how to replicate success patterns across teams.

The most meaningful measurement happens at the task level. Rather than asking "How many people logged in this month?" organizations should identify specific tasks where AI integration delivers measurable improvements. This approach helps articulate concrete outcomes and provides clearer guidance for expanding successful use cases.

Building a Culture of Experimentation

One of the strongest predictors of successful AI implementation is organizational encouragement of experimentation. However, creating this culture requires more than simply telling people to "try things." It demands psychological safety for failure, allocated time for exploration, and leadership that treats unsuccessful experiments as learning rather than waste.

Johnson & Johnson provides a compelling example. When they cut 80% of their AI projects, critics interpreted this as AI failure. However, leadership framed it differently: they were eliminating unsuccessful experiments to concentrate resources on the 20% delivering expected returns. This approach treats experimentation failures as valuable data rather than wasted investment.

Organizations can foster experimentation through hackathons, "promptathons," or dedicated exploration time. A particularly effective approach involves having people experiment with AI on problems they're not currently working on, encouraging creative exploration beyond immediate work pressures.

The Future of Change Management in an AI-Enabled Workplace

The conversation concluded with a vision for how AI might transform change management practice itself. Paul's concept of "personalized change management at scale" addresses a fundamental constraint in the field: the inability to provide individualized change support due to resource limitations.

Current change approaches typically segment by department or impacted group. AI could enable change practitioners to tailor communications, training, and support to individual needs, preferences, and concerns. Paul shared an example of automating leadership communications to highlight specific "What's In It For Me" messages relevant to his product team's priorities and concerns.

This represents the ultimate application of Prosci's principle that organizational change happens one person at a time. AI could help change practitioners get closer to individual needs while maintaining the efficiency required for enterprise-scale transformations.

Key Takeaways for Change Practitioners

The webinar highlighted several actionable insights for change practitioners supporting AI adoption:

1. Ensure leaders personally engage with AI tools rather than delegating to technical teams

2. Create safe spaces for experimentation while providing clear policy guidance

3. Focus on building change capability rather than just tool-specific training

4. Measure utilization and proficiency alongside adoption rates

5. Illuminate and leverage unsanctioned AI use happening in your organization

6. Apply frameworks like the AI Integration model to help people identify AI opportunities in their work

As AI continues evolving at unprecedented speed, the organizations that succeed will be those that recognize adoption as fundamentally a people challenge requiring structured change management approaches.