Recent polling shows that a meaningful share of workers, especially younger, well-educated knowledge workers, use artificial intelligence (AI) chatbots for core tasks like research, editing and drafting. With each update bringing new capabilities and broader adoption, this trend is projected to steadily increase.

At the same time, studies show that leaders are often unaware of how employees use chatbots in the workplace or underestimate adoption rates within their companies. That’s a problem. Without strategic oversight from learning and development (L&D) leaders, workplace chatbot usage might default to shallow or purely efficiency-driven approaches, that is, solving tasks quickly and superficially while overlooking deeper learning opportunities and professional growth.

Organizational AI culture is already forming in most companies. This culture can be understood as how employees, on a general level, use chatbots in their daily work. Crucially, it develops whether or not L&D specialists intervene. If they remain uninvolved, employees’ unguided generic chatbot usage will implicitly shape the organization’s internal AI culture, and it may not be an optimal one. Such a culture can dampen employee development and professional growth, ultimately limiting the company’s overall progress. For this reason, L&D specialists have a key role in intentionally shaping a human-centered AI culture that actively promotes workplace flourishing.

What is Human-Centered AI Culture?

Human-centered AI (HCAI) refers to AI use that actively supports human growth, satisfies core psychological needs and enhances performance in ways that facilitate rather than undermine those needs. When designed and guided well, HCAI can help employees leverage chatbots for deeper learning, skill building and professional development. Yet, decades of research on cognitive biases, cognitive load, workplace time pressure, stress and natural effort-avoidance show how easy it is to default to shallow, efficiency-only use, especially when employees lack awareness of strategies that make chatbot use genuinely developmental. Therefore, it should be clear that an HCAI culture in organizations will not emerge by accident. L&D teams must promote it intentionally and strategically.

How to Intentionally Shape a Human-Centered AI Culture

Step 1: Investigate

Investigate how chatbots are used in your organization by conducting internal surveys and short interviews across departments, gathering data on common chatbot use cases. Key areas to examine include: how employees currently use chatbots (formally and informally), which tasks are outsourced to chatbots, frequency of use and whether such usage contributes to skill building or undermines it. Depending on your organization’s work nature, you might want to adjust questions to obtain a broader picture.

Step 2: Assess

Analyze the results and compare them against a standard of human-centered AI use. A useful benchmark can be built from decades of research in organizational psychology, cognitive science and educational technology (see also Table 1).

For example, self-determination theory (SDT) identifies core human needs: mastery, autonomy and relatedness. You can analyze whether chatbot usage facilitates or stifles fulfillment of these needs. For instance, if someone uses chatbots to solve novel tasks instead of learning to do them independently, this may signal misalignment with the need for mastery (mastery is being outsourced). If someone becomes dependent on chatbots to accomplish tasks, the need for autonomy diminishes. Finally, if someone constantly substitutes live dialogue with colleagues for feedback or advice and relies solely on chatbots, this can undermine relatedness needs in the workplace.

Additionally, consider key factors that contribute to a workplace flourishing. Factors like purpose and meaning, creativity and innovative thinking, flow and engagement are crucial for professional development and growth (and indirectly influence organizational growth). Again, analyze whether AI usage patterns support or diminish these flourishing factors. For example, if someone routinely uses chatbots to solve high-order tasks by generating ready-made ideas, texts or pipelines, they may become disengaged, decreasing their sense of job meaning, which, if habitual, will diminish creative and innovative output.

Finally, educational technology research shows that effective learning, growth and innovation require preserving some degree of intellectual struggle. If accomplishing a task or learning a new skill isn’t challenging or slightly difficult, learning won’t be deep enough. Employees might use chatbots in ways that eliminate this struggle and growth component, which is an additional red flag indicating intervention is needed.

It’s not only a matter of employee development — all these collective chatbot use cases will ultimately shape your organization’s growth and value. Therefore, if you identify red flags when comparing your AI use culture with this benchmark, it might be the right time to intervene.

Table 1. An example benchmark for evaluating AI use culture.

Step 3: Intervene With Intentional Design

To mitigate the abovementioned concerns, L&D teams can embed HCAI practices into training sessions, onboarding and leadership development. They can also provide employees with practical AI usage guidelines, templates or prompt libraries designed for learning enhancement, and encourage peer-sharing of best practices. In short, there’s an opportunity to turn AI use into a collaborative learning activity.

Specifically, one could introduce the following practices:

1. Always create “raw material” first.

People are better protected in terms of cognitive development and creativity when they use chatbots to assess, critique and enhance their ideas rather than obtaining ideas from scratch. When using chatbots, always begin by drafting text, producing an initial idea, concept or pipeline, and only then use chatbots for critical evaluation. Chatbots can effectively provide feedback on ideas, critique them from multiple perspectives, reveal disadvantages and more. For example, before presenting an idea to a supervisor, workers could go through several rounds of critical revision with a chatbot. By creating the material first and then evaluating it with chatbots, employees preserve authorship, mastery and creativity and potentially enhance them further.

2. Prompt chatbots for exploration and insight.

While it’s tempting to use chatbots for ready-made answers, as shown above, this leads to disadvantages for both individuals and organizations. Therefore, L&D teams should train employees to create appropriate prompts for learning and development rather than cognitive outsourcing. Such prompts include asking chatbots not to provide direct solutions but to guide deeper exploration through provocative and insightful questions. Essentially, this approach uses chatbots as Socratic sparring partners that challenge ideas and investigate deeper assumptions and motivations.

3. Be skeptical.

Chatbots still have limitations and produce highly generic answers. Employees should be taught to question chatbot outputs. For example, when uncertain, they should ask: Where did the chatbot find this information? What are the implicit assumptions? Is there potentially a better or more creative solution? Is there a limitation in the training data?

4. Be creative.

Using chatbots in conventional ways leads to creative disadvantages. For example, research suggests that if all organizations use chatbots in the same way (ideation, brainstorming, rewrites, etc.), it’s very probable that over time their collective outputs would resemble each other, which means the company will not innovate. Hence, it’s better to create an idea first, then assess its novelty and creativity with a chatbot and refine it based on feedback. However, an additional strategy is using chatbots in unconventional, creative ways. For instance, employees can prompt chatbots to act like a personal advisory board consisting of different personas with completely different viewpoints, rewrite their ideas from a controversial perspective or from the perspective of various industries.

5. Alternate AI and human processes.

Collaborative human discussions often generate unexpected connections that emerge from the dynamic interplay of heterogeneous backgrounds. This is something that’s difficult to achieve in isolated AI interactions. Regular human check-ins also serve as a reality test, ensuring that AI-assisted work remains grounded in practical constraints and aligned with team goals. Further, human colleagues can bring contextual knowledge, emotional intelligence and diverse lived experiences that AI cannot replicate. They can challenge assumptions based on real-world implementation experience and offer insights rooted in organizational culture and politics.

Final Thoughts

Various surveys on the future of work show that in the coming years and decades, not only technical skills will be valued, but also human skills that are difficult to substitute with technology alone. Creativity, innovation, critical thinking and emotional intelligence will be crucial workplace qualities where technical skills (with AI’s help) might become commoditized. However, default efficiency-driven chatbot usage risks undermining these exact qualities. To effectively prepare human resources for the future, L&D specialists and other organizational leaders must intentionally shape corporate AI use and align it with human-centered principles. Without such engagement, the rise of AI use within your culture may lead to a decline in the skills your team needs to thrive in the future.